Abstract
Novel test selectors have demonstrated their effectiveness in accelerating the closure of functional coverage for various industrial digital designs in simulation-based verification. The primary advantages of these test selectors include performance that is not impacted by coverage holes, straightforward implementation, and relatively low computational expense. However, the detection of stimuli with novel temporal patterns remains largely unexplored. This paper introduces two novel test selectors designed to identify such stimuli. The experiments reveal that both test selectors can accelerate the functional coverage for a commercial bus bridge, compared to random test selection. Specifically, one selector achieves a 26.9\% reduction in the number of simulated tests required to reach 98.5\% coverage, outperforming the savings achieved by two previously published test selectors by factors of 13 and 2.68, respectively.
Title:
RaCIL: Ray Tracing based Multi-UAV Obstacle Avoidance through Composite Imitation Learning
Abstract
In this study, we address the challenge of obstacle avoidance for Unmanned Aerial Vehicles (UAVs) through an innovative composite imitation learning approach that combines Proximal Policy Optimization (PPO) with Behavior Cloning (BC) and Generative Adversarial Imitation Learning (GAIL), enriched by the integration of ray-tracing techniques. Our research underscores the significant role of ray-tracing in enhancing obstacle detection and avoidance capabilities. Moreover, we demonstrate the effectiveness of incorporating GAIL in coordinating the flight paths of two UAVs, showcasing improved collision avoidance capabilities. Extending our methodology, we apply our combined PPO, BC, GAIL, and ray-tracing framework to scenarios involving four UAVs, illustrating its scalability and adaptability to more complex scenarios. The findings indicate that our approach not only improves the reliability of basic PPO based obstacle avoidance but also paves the way for advanced autonomous UAV operations in crowded or dynamic environments.
Title:
Robust ADAS: Enhancing Robustness of Machine Learning-based Advanced Driver Assistance Systems for Adverse Weather
Authors: Muhammad Zaeem Shahzad, Muhammad Abdullah Hanif, Muhammad Shafique
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
In the realm of deploying Machine Learning-based Advanced Driver Assistance Systems (ML-ADAS) into real-world scenarios, adverse weather conditions pose a significant challenge. Conventional ML models trained on clear weather data falter when faced with scenarios like extreme fog or heavy rain, potentially leading to accidents and safety hazards. This paper addresses this issue by proposing a novel approach: employing a Denoising Deep Neural Network as a preprocessing step to transform adverse weather images into clear weather images, thereby enhancing the robustness of ML-ADAS systems. The proposed method eliminates the need for retraining all subsequent Depp Neural Networks (DNN) in the ML-ADAS pipeline, thus saving computational resources and time. Moreover, it improves driver visualization, which is critical for safe navigation in adverse weather conditions. By leveraging the UNet architecture trained on an augmented KITTI dataset with synthetic adverse weather images, we develop the Weather UNet (WUNet) DNN to remove weather artifacts. Our study demonstrates substantial performance improvements in object detection with WUNet preprocessing under adverse weather conditions. Notably, in scenarios involving extreme fog, our proposed solution improves the mean Average Precision (mAP) score of the YOLOv8n from 4% to 70%.
Title:
Novel Human Machine Interface via Robust Hand Gesture Recognition System using Channel Pruned YOLOv5s Model
Authors: Abir Sen, Tapas Kumar Mishra, Ratnakar Dash
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Hand gesture recognition (HGR) is a vital component in enhancing the human-computer interaction experience, particularly in multimedia applications, such as virtual reality, gaming, smart home automation systems, etc. Users can control and navigate through these applications seamlessly by accurately detecting and recognizing gestures. However, in a real-time scenario, the performance of the gesture recognition system is sometimes affected due to the presence of complex background, low-light illumination, occlusion problems, etc. Another issue is building a fast and robust gesture-controlled human-computer interface (HCI) in the real-time scenario. The overall objective of this paper is to develop an efficient hand gesture detection and classification model using a channel-pruned YOLOv5-small model and utilize the model to build a gesture-controlled HCI with a quick response time (in ms) and higher detection speed (in fps). First, the YOLOv5s model is chosen for the gesture detection task. Next, the model is simplified by using a channel-pruned algorithm. After that, the pruned model is further fine-tuned to ensure detection efficiency. We have compared our suggested scheme with other state-of-the-art works, and it is observed that our model has shown superior results in terms of mAP (mean average precision), precision (\%), recall (\%), and F1-score (\%), fast inference time (in ms), and detection speed (in fps). Our proposed method paves the way for deploying a pruned YOLOv5s model for a real-time gesture-command-based HCI to control some applications, such as the VLC media player, Spotify player, etc., using correctly classified gesture commands in real-time scenarios. The average detection speed of our proposed system has reached more than 60 frames per second (fps) in real-time, which meets the perfect requirement in real-time application control.
Title:
Ensuring Responsible Sourcing of Large Language Model Training Data Through Knowledge Graph Comparison
Authors: Devam Mondal, Carlo Lipizzi
Subjects: Subjects:
Computation and Language (cs.CL); Machine Learning (cs.LG)
Abstract
In light of recent plagiarism allegations Brough by publishers, newspapers, and other creators of copyrighted corpora against large language model (LLM) developers, we propose a novel system, a variant of a plagiarism detection system, that assesses whether a knowledge source has been used in the training or fine-tuning of a large language model. Unlike current methods, we utilize an approach that uses Resource Description Framework (RDF) triples to create knowledge graphs from both a source document and a LLM continuation of that document. These graphs are then analyzed with respect to content using cosine similarity and with respect to structure using a normalized version of graph edit distance that shows the degree of isomorphism. Unlike traditional systems that focus on content matching and keyword identification between a source and target corpus, our approach enables a broader evaluation of similarity and thus a more accurate comparison of the similarity between a source document and LLM continuation by focusing on relationships between ideas and their organization with regards to others. Additionally, our approach does not require access to LLM metrics like perplexity that may be unavailable in closed large language modeling "black-box" systems, as well as the training corpus. A prototype of our system will be found on a hyperlinked GitHub repository.
Title:
SMILe: Leveraging Submodular Mutual Information For Robust Few-Shot Object Detection
Authors: Anay Majee, Ryan Sharp, Rishabh Iyer
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Confusion and forgetting of object classes have been challenges of prime interest in Few-Shot Object Detection (FSOD). To overcome these pitfalls in metric learning based FSOD techniques, we introduce a novel Submodular Mutual Information Learning (SMILe) framework which adopts combinatorial mutual information functions to enforce the creation of tighter and discriminative feature clusters in FSOD. Our proposed approach generalizes to several existing approaches in FSOD, agnostic of the backbone architecture demonstrating elevated performance gains. A paradigm shift from instance based objective functions to combinatorial objectives in SMILe naturally preserves the diversity within an object class resulting in reduced forgetting when subjected to few training examples. Furthermore, the application of mutual information between the already learnt (base) and newly added (novel) objects ensures sufficient separation between base and novel classes, minimizing the effect of class confusion. Experiments on popular FSOD benchmarks, PASCAL-VOC and MS-COCO show that our approach generalizes to State-of-the-Art (SoTA) approaches improving their novel class performance by up to 5.7% (3.3 mAP points) and 5.4% (2.6 mAP points) on the 10-shot setting of VOC (split 3) and 30-shot setting of COCO datasets respectively. Our experiments also demonstrate better retention of base class performance and up to 2x faster convergence over existing approaches agnostic of the underlying architecture.
Title:
Adversarial Magnification to Deceive Deepfake Detection through Super Resolution
Abstract
Deepfake technology is rapidly advancing, posing significant challenges to the detection of manipulated media content. Parallel to that, some adversarial attack techniques have been developed to fool the deepfake detectors and make deepfakes even more difficult to be detected. This paper explores the application of super resolution techniques as a possible adversarial attack in deepfake detection. Through our experiments, we demonstrate that minimal changes made by these methods in the visual appearance of images can have a profound impact on the performance of deepfake detection systems. We propose a novel attack using super resolution as a quick, black-box and effective method to camouflage fake images and/or generate false alarms on pristine images. Our results indicate that the usage of super resolution can significantly impair the accuracy of deepfake detectors, thereby highlighting the vulnerability of such systems to adversarial attacks. The code to reproduce our experiments is available at: this https URL
Title:
Enhancing Scalability of Optimal Kron-based Reduction of Networks (Opti-KRON) via Decomposition with Community Detection
Authors: Omid Mokhtari, Samuel Chevalier, Mads Almassalkhi
Abstract
Electrical networks contain thousands of interconnected nodes and edges, which leads to computational challenges in some power system studies. To address these challenges, we contend that network reductions can serve as a framework to enable scalable computing in power systems. By building upon a prior AC "Opti-KRON" formulation, this paper presents a DC power flow formulation for finding network reductions that are optimal within the context of large transmission analysis. Opti-KRON previously formulated optimal Kron-based network reductions as a mixed integer linear program (MILP), where the number of binary variables scaled with the number of nodes. To improve the scalability of the Opti-KRON approach, we augment the MILP formulation with a community detection (CD) technique that segments a large network into smaller, disjoint, but contiguous sub-graphs (i.e., communities). For each sub-graph, we then (in parallel) apply MILP-based along with a new cutting plane constraint, thus, enhancing scalability. Ultimately, the new DC-based Opti-KRON method can achieve a 80-95\% reduction of networks (in terms of nodes) while statistically outperforming other CD- and Kron-based methods. We present simulation results for the IEEE RTS-96 and the 2383-bus Polish networks.
Title:
Generalized Event Cameras
Authors: Varun Sundar, Matthew Dutson, Andrei Ardelean, Claudio Bruschini, Edoardo Charbon, Mohit Gupta
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV)
Abstract
Event cameras capture the world at high time resolution and with minimal bandwidth requirements. However, event streams, which only encode changes in brightness, do not contain sufficient scene information to support a wide variety of downstream tasks. In this work, we design generalized event cameras that inherently preserve scene intensity in a bandwidth-efficient manner. We generalize event cameras in terms of when an event is generated and what information is transmitted. To implement our designs, we turn to single-photon sensors that provide digital access to individual photon detections; this modality gives us the flexibility to realize a rich space of generalized event cameras. Our single-photon event cameras are capable of high-speed, high-fidelity imaging at low readout rates. Consequently, these event cameras can support plug-and-play downstream inference, without capturing new event datasets or designing specialized event-vision models. As a practical implication, our designs, which involve lightweight and near-sensor-compatible computations, provide a way to use single-photon sensors without exorbitant bandwidth costs.
Title:
Navigating Connected Car Cybersecurity: Location Anomaly Detection with RAN Data
Authors: Feng Wang, Yaron Koral, Kenichi Futamura
Subjects: Subjects:
Cryptography and Security (cs.CR); Networking and Internet Architecture (cs.NI)
Abstract
The cybersecurity of connected cars, integral to the broader Internet of Things (IoT) landscape, has become of paramount concern. Cyber-attacks, including hijacking and spoofing, pose significant threats to these technological advancements, potentially leading to unauthorized control over vehicular networks or creating deceptive identities. Given the difficulty of deploying comprehensive defensive logic across all vehicles, this paper presents a novel approach for identifying potential attacks through Radio Access Network (RAN) event monitoring. The major contribution of this paper is a location anomaly detection module that identifies aberrant devices that appear in multiple locations simultaneously - a potential indicator of a hijacking attack. We demonstrate how RAN-event based location anomaly detection is effective in combating malicious activity targeting connected cars. Using RAN data generated by tens of millions of connected cars, we developed a fast and efficient method for identifying potential malicious or rogue devices. The implications of this research are far-reaching. By increasing the security of connected cars, we can enhance the safety of users, provide robust defenses for the automotive industry, and improve overall cybersecurity practices for IoT devices.
Title:
ADFQ-ViT: Activation-Distribution-Friendly Post-Training Quantization for Vision Transformers
Authors: Yanfeng Jiang, Ning Sun, Xueshuo Xie, Fei Yang, Tao Li
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Vision Transformers (ViTs) have exhibited exceptional performance across diverse computer vision tasks, while their substantial parameter size incurs significantly increased memory and computational demands, impeding effective inference on resource-constrained devices. Quantization has emerged as a promising solution to mitigate these challenges, yet existing methods still suffer from significant accuracy loss at low-bit. We attribute this issue to the distinctive distributions of post-LayerNorm and post-GELU activations within ViTs, rendering conventional hardware-friendly quantizers ineffective, particularly in low-bit scenarios. To address this issue, we propose a novel framework called Activation-Distribution-Friendly post-training Quantization for Vision Transformers, ADFQ-ViT. Concretely, we introduce the Per-Patch Outlier-aware Quantizer to tackle irregular outliers in post-LayerNorm activations. This quantizer refines the granularity of the uniform quantizer to a per-patch level while retaining a minimal subset of values exceeding a threshold at full-precision. To handle the non-uniform distributions of post-GELU activations between positive and negative regions, we design the Shift-Log2 Quantizer, which shifts all elements to the positive region and then applies log2 quantization. Moreover, we present the Attention-score enhanced Module-wise Optimization which adjusts the parameters of each quantizer by reconstructing errors to further mitigate quantization error. Extensive experiments demonstrate ADFQ-ViT provides significant improvements over various baselines in image classification, object detection, and instance segmentation tasks at 4-bit. Specifically, when quantizing the ViT-B model to 4-bit, we achieve a 10.23% improvement in Top-1 accuracy on the ImageNet dataset.
Title:
Efficient DNN-Powered Software with Fair Sparse Models
Abstract
With the emergence of the Software 3.0 era, there is a growing trend of compressing and integrating large models into software systems, with significant societal implications. Regrettably, in numerous instances, model compression techniques impact the fairness performance of these models and thus the ethical behavior of DNN-powered software. One of the most notable example is the Lottery Ticket Hypothesis (LTH), a prevailing model pruning approach. This paper demonstrates that fairness issue of LTHbased pruning arises from both its subnetwork selection and training procedures, highlighting the inadequacy of existing remedies. To address this, we propose a novel pruning framework, Ballot, which employs a novel conflict-detection-based subnetwork selection to find accurate and fair subnetworks, coupled with a refined training process to attain a high-performance model, thereby improving the fairness of DNN-powered software. By means of this procedure, Ballot improves the fairness of pruning by 38.00%, 33.91%, 17.96%, and 35.82% compared to state-of-the-art baselines, namely Magnitude Pruning, Standard LTH, SafeCompress, and FairScratch respectively, based on our evaluation of five popular datasets and three widely used models. Our code is available at https://anonymous.4open.science/r/Ballot-506E.
Title:
Investigating the Contextualised Word Embedding Dimensions Responsible for Contextual and Temporal Semantic Changes
Authors: Taichi Aida, Danushka Bollegala
Subjects: Subjects:
Computation and Language (cs.CL)
Abstract
Words change their meaning over time as well as in different contexts. The sense-aware contextualised word embeddings (SCWEs) such as the ones produced by XL-LEXEME by fine-tuning masked langauge models (MLMs) on Word-in-Context (WiC) data attempt to encode such semantic changes of words within the contextualised word embedding (CWE) spaces. Despite the superior performance of SCWEs in contextual/temporal semantic change detection (SCD) benchmarks, it remains unclear as to how the meaning changes are encoded in the embedding space. To study this, we compare pre-trained CWEs and their fine-tuned versions on contextual and temporal semantic change benchmarks under Principal Component Analysis (PCA) and Independent Component Analysis (ICA) transformations. Our experimental results reveal several novel insights such as (a) although there exist a smaller number of axes that are responsible for semantic changes of words in the pre-trained CWE space, this information gets distributed across all dimensions when fine-tuned, and (b) in contrast to prior work studying the geometry of CWEs, we find that PCA to better represent semantic changes than ICA. Source code is available at this https URL .
Title:
A Radiometric Correction based Optical Modeling Approach to Removing Reflection Noise in TLS Point Clouds of Urban Scenes
Authors: Li Fang, Tianyu Li, Yanghong Lin, Shudong Zhou, Wei Yao
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV)
Abstract
Point clouds are vital in computer vision tasks such as 3D reconstruction, autonomous driving, and robotics. However, TLS-acquired point clouds often contain virtual points from reflective surfaces, causing disruptions. This study presents a reflection noise elimination algorithm for TLS point clouds. Our innovative reflection plane detection algorithm, based on geometry-optical models and physical properties, identifies and categorizes reflection points per optical reflection theory. We've adapted the LSFH feature descriptor to retain reflection features, mitigating interference from symmetrical architectural structures. By incorporating the Hausdorff feature distance, the algorithm enhances resilience to ghosting and deformation, improving virtual point detection accuracy. Extensive experiments on the 3DRN benchmark dataset, featuring diverse urban environments with virtual TLS reflection noise, show our algorithm improves precision and recall rates for 3D points in reflective regions by 57.03\% and 31.80\%, respectively. Our method achieves a 9.17\% better outlier detection rate and 5.65\% higher accuracy than leading methods. Access the 3DRN dataset at (this https URL).
Title:
A Pairwise DomMix Attentive Adversarial Network for Unsupervised Domain Adaptive Object Detection
Authors: Jie Shao, Jiacheng Wu, Wenzhong Shen, Cheng Yang
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Unsupervised Domain Adaptive Object Detection (DAOD) could adapt a model trained on a source domain to an unlabeled target domain for object detection. Existing unsupervised DAOD methods usually perform feature alignments from the target to the source. Unidirectional domain transfer would omit information about the target samples and result in suboptimal adaptation when there are large domain shifts. Therefore, we propose a pairwise attentive adversarial network with a Domain Mixup (DomMix) module to mitigate the aforementioned challenges. Specifically, a deep-level mixup is employed to construct an intermediate domain that allows features from both domains to share their differences. Then a pairwise attentive adversarial network is applied with attentive encoding on both image-level and instance-level features at different scales and optimizes domain alignment by adversarial learning. This allows the network to focus on regions with disparate contextual information and learn their similarities between different domains. Extensive experiments are conducted on several benchmark datasets, demonstrating the superiority of our proposed method.
Title:
FedPot: A Quality-Aware Collaborative and Incentivized Honeypot-Based Detector for Smart Grid Networks
Authors: Abdullatif Albaseer, Nima Abdi, Mohamed Abdallah, Marwa Qaraqe, Saif Alkuwari
Subjects: Subjects:
Networking and Internet Architecture (cs.NI)
Abstract
Honeypot technologies provide an effective defense strategy for the Industrial Internet of Things (IIoT), particularly in enhancing the Advanced Metering Infrastructure's (AMI) security by bolstering the network intrusion detection system. For this security paradigm to be fully realized, it necessitates the active participation of small-scale power suppliers (SPSs) in implementing honeypots and engaging in collaborative data sharing with traditional power retailers (TPRs). To motivate this interaction, TPRs incentivize data sharing with tangible rewards. However, without access to an SPS's confidential data, it is daunting for TPRs to validate shared data, thereby risking SPSs' privacy and increasing sharing costs due to voluminous honeypot logs. These challenges can be resolved by utilizing Federated Learning (FL), a distributed machine learning (ML) technique that allows for model training without data relocation. However, the conventional FL algorithm lacks the requisite functionality for both the security defense model and the rewards system of the AMI network. This work presents two solutions: first, an enhanced and cost-efficient FedAvg algorithm incorporating a novel data quality measure, and second, FedPot, the development of an effective security model with a fair incentives mechanism under an FL architecture. Accordingly, SPSs are limited to sharing the ML model they learn after efficiently measuring their local data quality, whereas TPRs can verify the participants' uploaded models and fairly compensate each participant for their contributions through rewards. Simulation results, drawn from realistic mircorgrid network log datasets, demonstrate that the proposed solutions outperform state-of-the-art techniques by enhancing the security model and guaranteeing fair reward distributions.
Title:
Early-Stage Anomaly Detection: A Study of Model Performance on Complete vs. Partial Flows
Authors: Adrian Pekar, Richard Jozsa
Subjects: Subjects:
Machine Learning (cs.LG); Cryptography and Security (cs.CR)
Abstract
This study investigates the efficacy of machine learning models, specifically Random Forest, in anomaly detection systems when trained on complete flow records and tested on partial flow data. We explore the performance disparity that arises when models are applied to incomplete data typical in real-world, real-time network environments. Our findings demonstrate a significant decline in model performance, with precision and recall dropping by up to 30\% under certain conditions when models trained on complete flows are tested against partial flows. Conversely, models trained and tested on consistently complete or partial datasets maintain robustness, highlighting the importance of dataset consistency in training. The study reveals that a minimum of 7 packets in the test set is required for maintaining reliable detection rates. These results underscore the need for tailored training strategies that can effectively adapt to the dynamics of partial data, enhancing the practical applicability of anomaly detection systems in operational settings.
Title:
A Self-Supervised Task for Fault Detection in Satellite Multivariate Time Series
Authors: Carlo Cena, Silvia Bucci, Alessandro Balossino, Marcello Chiaberge
Abstract
In the space sector, due to environmental conditions and restricted accessibility, robust fault detection methods are imperative for ensuring mission success and safeguarding valuable assets. This work proposes a novel approach leveraging Physics-Informed Real NVP neural networks, renowned for their ability to model complex and high-dimensional distributions, augmented with a self-supervised task based on sensors' data permutation. It focuses on enhancing fault detection within the satellite multivariate time series. The experiments involve various configurations, including pre-training with self-supervision, multi-task learning, and standalone self-supervised training. Results indicate significant performance improvements across all settings. In particular, employing only the self-supervised loss yields the best overall results, suggesting its efficacy in guiding the network to extract relevant features for fault detection. This study presents a promising direction for improving fault detection in space systems and warrants further exploration in other datasets and applications.
Title:
Robot Shape and Location Retention in Video Generation Using Diffusion Models
Abstract
Diffusion models have marked a significant milestone in the enhancement of image and video generation technologies. However, generating videos that precisely retain the shape and location of moving objects such as robots remains a challenge. This paper presents diffusion models specifically tailored to generate videos that accurately maintain the shape and location of mobile robots. This development offers substantial benefits to those working on detecting dangerous interactions between humans and robots by facilitating the creation of training data for collision detection models, circumventing the need for collecting data from the real world, which often involves legal and ethical issues. Our models incorporate techniques such as embedding accessible robot pose information and applying semantic mask regulation within the ConvNext backbone network. These techniques are designed to refine intermediate outputs, therefore improving the retention performance of shape and location. Through extensive experimentation, our models have demonstrated notable improvements in maintaining the shape and location of different robots, as well as enhancing overall video generation quality, compared to the benchmark diffusion model. Codes will be opensourced at \href{this https URL}{Github}.
Title:
Complex Event Recognition with Symbolic Register Transducers: Extended Technical Report
Authors: Elias Alevizos, Alexander Artikis, Georgios Paliouras
Subjects: Subjects:
Formal Languages and Automata Theory (cs.FL); Artificial Intelligence (cs.AI); Databases (cs.DB)
Abstract
We present a system for Complex Event Recognition (CER) based on automata. While multiple such systems have been described in the literature, they typically suffer from a lack of clear and denotational semantics, a limitation which often leads to confusion with respect to their expressive power. In order to address this issue, our system is based on an automaton model which is a combination of symbolic and register automata. We extend previous work on these types of automata, in order to construct a formalism with clear semantics and a corresponding automaton model whose properties can be formally investigated. We call such automata Symbolic Register Transducers (SRT). We show that SRT are closed under various operators, but are not in general closed under complement and they are not determinizable. However, they are closed under these operations when a window operator, quintessential in Complex Event Recognition, is used. We show how SRT can be used in CER in order to detect patterns upon streams of events, using our framework that provides declarative and compositional semantics, and that allows for a systematic treatment of such automata. For SRT to work in pattern detection, we allow them to mark events from the input stream as belonging to a complex event or not, hence the name "transducers". We also present an implementation of SRT which can perform CER. We compare our SRT-based CER engine against other state-of-the-art CER systems and show that it is both more expressive and more efficient.
Title:
A Wolf in Sheep's Clothing: Practical Black-box Adversarial Attacks for Evading Learning-based Windows Malware Detection in the Wild
Authors: Xiang Ling, Zhiyu Wu, Bin Wang, Wei Deng, Jingzheng Wu, Shouling Ji, Tianyue Luo, Yanjun Wu
Subjects: Subjects:
Cryptography and Security (cs.CR)
Abstract
Given the remarkable achievements of existing learning-based malware detection in both academia and industry, this paper presents MalGuise, a practical black-box adversarial attack framework that evaluates the security risks of existing learning-based Windows malware detection systems under the black-box setting. MalGuise first employs a novel semantics-preserving transformation of call-based redividing to concurrently manipulate both nodes and edges of malware's control-flow graph, making it less noticeable. By employing a Monte-Carlo-tree-search-based optimization, MalGuise then searches for an optimized sequence of call-based redividing transformations to apply to the input Windows malware for evasions. Finally, it reconstructs the adversarial malware file based on the optimized transformation sequence while adhering to Windows executable format constraints, thereby maintaining the same semantics as the original. MalGuise is systematically evaluated against three state-of-the-art learning-based Windows malware detection systems under the black-box setting. Evaluation results demonstrate that MalGuise achieves a remarkably high attack success rate, mostly exceeding 95%, with over 91% of the generated adversarial malware files maintaining the same semantics. Furthermore, MalGuise achieves up to a 74.97% attack success rate against five anti-virus products, highlighting potential tangible security concerns to real-world users.
Title:
Domain-independent detection of known anomalies
Authors: Jonas Bühler, Jonas Fehrenbach, Lucas Steinmann, Christian Nauck, Marios Koulakis
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
One persistent obstacle in industrial quality inspection is the detection of anomalies. In real-world use cases, two problems must be addressed: anomalous data is sparse and the same types of anomalies need to be detected on previously unseen objects. Current anomaly detection approaches can be trained with sparse nominal data, whereas domain generalization approaches enable detecting objects in previously unseen domains. Utilizing those two observations, we introduce the hybrid task of domain generalization on sparse classes. To introduce an accompanying dataset for this task, we present a modification of the well-established MVTec AD dataset by generating three new datasets. In addition to applying existing methods for benchmark, we design two embedding-based approaches, Spatial Embedding MLP (SEMLP) and Labeled PatchCore. Overall, SEMLP achieves the best performance with an average image-level AUROC of 87.2 % vs. 80.4 % by MIRO. The new and openly available datasets allow for further research to improve industrial anomaly detection.
Title:
Efficient IoT Devices Localization Through Wi-Fi CSI Feature Fusion and Anomaly Detection
Authors: Yan Li, Jie Yang, Shang-Ling Shih, Wan-Ting Shih, Chao-Kai Wen, Shi Jin
Abstract
Internet of Things (IoT) device localization is fundamental to smart home functionalities, including indoor navigation and tracking of individuals. Traditional localization relies on relative methods utilizing the positions of anchors within a home environment, yet struggles with precision due to inherent inaccuracies in these anchor positions. In response, we introduce a cutting-edge smartphone-based localization system for IoT devices, leveraging the precise positioning capabilities of smartphones equipped with motion sensors. Our system employs artificial intelligence (AI) to merge channel state information from proximal trajectory points of a single smartphone, significantly enhancing line of sight (LoS) angle of arrival (AoA) estimation accuracy, particularly under severe multipath conditions. Additionally, we have developed an AI-based anomaly detection algorithm to further increase the reliability of LoSAoA estimation. This algorithm improves measurement reliability by analyzing the correlation between the accuracy of reversed feature reconstruction and the LoS-AoA estimation. Utilizing a straightforward least squares algorithm in conjunction with accurate LoS-AoA estimation and smartphone positional data, our system efficiently identifies IoT device locations. Validated through extensive simulations and experimental tests with a receiving antenna array comprising just two patch antenna elements in the horizontal direction, our methodology has been shown to attain decimeter-level localization accuracy in nearly 90% of cases, demonstrating robust performance even in challenging real-world scenarios. Additionally, our proposed anomaly detection algorithm trained on Wi-Fi data can be directly applied to ultra-wideband, also outperforming the most advanced techniques.
Title:
Unified Anomaly Detection methods on Edge Device using Knowledge Distillation and Quantization
Abstract
With the rapid advances in deep learning and smart manufacturing in Industry 4.0, there is an imperative for high-throughput, high-performance, and fully integrated visual inspection systems. Most anomaly detection approaches using defect detection datasets, such as MVTec AD, employ one-class models that require fitting separate models for each class. On the contrary, unified models eliminate the need for fitting separate models for each class and significantly reduce cost and memory requirements. Thus, in this work, we experiment with considering a unified multi-class setup. Our experimental study shows that multi-class models perform at par with one-class models for the standard MVTec AD dataset. Hence, this indicates that there may not be a need to learn separate object/class-wise models when the object classes are significantly different from each other, as is the case of the dataset considered. Furthermore, we have deployed three different unified lightweight architectures on the CPU and an edge device (NVIDIA Jetson Xavier NX). We analyze the quantized multi-class anomaly detection models in terms of latency and memory requirements for deployment on the edge device while comparing quantization-aware training (QAT) and post-training quantization (PTQ) for performance at different precision widths. In addition, we explored two different methods of calibration required in post-training scenarios and show that one of them performs notably better, highlighting its importance for unsupervised tasks. Due to quantization, the performance drop in PTQ is further compensated by QAT, which yields at par performance with the original 32-bit Floating point in two of the models considered.
Title:
Zero-X: A Blockchain-Enabled Open-Set Federated Learning Framework for Zero-Day Attack Detection in IoV
Abstract
The Internet of Vehicles (IoV) is a crucial technology for Intelligent Transportation Systems (ITS) that integrates vehicles with the Internet and other entities. The emergence of 5G and the forthcoming 6G networks presents an enormous potential to transform the IoV by enabling ultra-reliable, low-latency, and high-bandwidth communications. Nevertheless, as connectivity expands, cybersecurity threats have become a significant concern. The issue has been further exacerbated by the rising number of zero-day (0-day) attacks, which can exploit unknown vulnerabilities and bypass existing Intrusion Detection Systems (IDSs). In this paper, we propose Zero-X, an innovative security framework that effectively detects both 0-day and N-day attacks. The framework achieves this by combining deep neural networks with Open-Set Recognition (OSR). Our approach introduces a novel scheme that uses blockchain technology to facilitate trusted and decentralized federated learning (FL) of the ZeroX framework. This scheme also prioritizes privacy preservation, enabling both CAVs and Security Operation Centers (SOCs) to contribute their unique knowledge while protecting the privacy of their sensitive data. To the best of our knowledge, this is the first work to leverage OSR in combination with privacy-preserving FL to identify both 0-day and N-day attacks in the realm of IoV. The in-depth experiments on two recent network traffic datasets show that the proposed framework achieved a high detection rate while minimizing the false positive rate. Comparison with related work showed that the Zero-X framework outperforms existing solutions.
Title:
Mast Kalandar at SemEval-2024 Task 8: On the Trail of Textual Origins: RoBERTa-BiLSTM Approach to Detect AI-Generated Text
Abstract
Large Language Models (LLMs) have showcased impressive abilities in generating fluent responses to diverse user queries. However, concerns regarding the potential misuse of such texts in journalism, educational, and academic contexts have surfaced. SemEval 2024 introduces the task of Multigenerator, Multidomain, and Multilingual Black-Box Machine-Generated Text Detection, aiming to develop automated systems for identifying machine-generated text and detecting potential misuse. In this paper, we i) propose a RoBERTa-BiLSTM based classifier designed to classify text into two categories: AI-generated or human ii) conduct a comparative study of our model with baseline approaches to evaluate its effectiveness. This paper contributes to the advancement of automatic text detection systems in addressing the challenges posed by machine-generated text misuse. Our architecture ranked 46th on the official leaderboard with an accuracy of 80.83 among 125.
Title:
YOLOv5, YOLOv8 and YOLOv10: The Go-To Detectors for Real-time Vision
Authors: Muhammad Hussain
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
This paper presents a comprehensive review of the evolution of the YOLO (You Only Look Once) object detection algorithm, focusing on YOLOv5, YOLOv8, and YOLOv10. We analyze the architectural advancements, performance improvements, and suitability for edge deployment across these versions. YOLOv5 introduced significant innovations such as the CSPDarknet backbone and Mosaic Augmentation, balancing speed and accuracy. YOLOv8 built upon this foundation with enhanced feature extraction and anchor-free detection, improving versatility and performance. YOLOv10 represents a leap forward with NMS-free training, spatial-channel decoupled downsampling, and large-kernel convolutions, achieving state-of-the-art performance with reduced computational overhead. Our findings highlight the progressive enhancements in accuracy, efficiency, and real-time performance, particularly emphasizing their applicability in resource-constrained environments. This review provides insights into the trade-offs between model complexity and detection accuracy, offering guidance for selecting the most appropriate YOLO version for specific edge computing applications.
Title:
Vision-driven Automated Mobile GUI Testing via Multimodal Large Language Model
Authors: Zhe Liu, Cheng Li, Chunyang Chen, Junjie Wang, Boyu Wu, Yawen Wang, Jun Hu, Qing Wang
Abstract
With the advancement of software rendering techniques, GUI pages in mobile apps now encompass a wealth of visual information, where the visual semantics of each page contribute to the overall app logic, presenting new challenges to software testing. Despite the progress in automated Graphical User Interface (GUI) testing, the absence of testing oracles has constrained its efficacy to identify only crash bugs with evident abnormal signals. Nonetheless, there are still a considerable number of non-crash bugs, ranging from unexpected behaviors to misalignments, often evading detection by existing techniques. While these bugs can exhibit visual cues that serve as potential testing oracles, they often entail a sequence of screenshots, and detecting them necessitates an understanding of the operational logic among GUI page transitions, which is challenging traditional techniques. Considering the remarkable performance of Multimodal Large Language Models (MLLM) in visual and language understanding, this paper proposes a vision-driven automated GUI testing approach VisionDroid to detect non-crash functional bugs with MLLM. It begins by extracting GUI text information and aligning it with screenshots to form a vision prompt, enabling MLLM to understand GUI context. The function-aware explorer then employs MLLM for deeper and function-oriented GUI page exploration, while the logic-aware bug detector segments the entire exploration history into logically cohesive parts and prompts the MLLM for bug detection. We evaluate VisionDroid on three datasets and compare it with 10 baselines, demonstrating its excellent performance. The ablation study further proves the contribution of each module. Moreover, VisionDroid identifies 29 new bugs on Google Play, of which 19 have been confirmed and fixed.
Title:
Enhancements for Real-Time Monte-Carlo Tree Search in General Video Game Playing
Authors: Dennis J.N.J. Soemers, Chiara F. Sironi, Torsten Schuster, Mark H.M. Winands
Abstract
General Video Game Playing (GVGP) is a field of Artificial Intelligence where agents play a variety of real-time video games that are unknown in advance. This limits the use of domain-specific heuristics. Monte-Carlo Tree Search (MCTS) is a search technique for game playing that does not rely on domain-specific knowledge. This paper discusses eight enhancements for MCTS in GVGP; Progressive History, N-Gram Selection Technique, Tree Reuse, Breadth-First Tree Initialization, Loss Avoidance, Novelty-Based Pruning, Knowledge-Based Evaluations, and Deterministic Game Detection. Some of these are known from existing literature, and are either extended or introduced in the context of GVGP, and some are novel enhancements for MCTS. Most enhancements are shown to provide statistically significant increases in win percentages when applied individually. When combined, they increase the average win percentage over sixty different games from 31.0% to 48.4% in comparison to a vanilla MCTS implementation, approaching a level that is competitive with the best agents of the GVG-AI competition in 2015.
Title:
Federated Learning for Zero-Day Attack Detection in 5G and Beyond V2X Networks
Abstract
Deploying Connected and Automated Vehicles (CAVs) on top of 5G and Beyond networks (5GB) makes them vulnerable to increasing vectors of security and privacy attacks. In this context, a wide range of advanced machine/deep learning based solutions have been designed to accurately detect security attacks. Specifically, supervised learning techniques have been widely applied to train attack detection models. However, the main limitation of such solutions is their inability to detect attacks different from those seen during the training phase, or new attacks, also called zero-day attacks. Moreover, training the detection model requires significant data collection and labeling, which increases the communication overhead, and raises privacy concerns. To address the aforementioned limits, we propose in this paper a novel detection mechanism that leverages the ability of the deep auto-encoder method to detect attacks relying only on the benign network traffic pattern. Using federated learning, the proposed intrusion detection system can be trained with large and diverse benign network traffic, while preserving the CAVs privacy, and minimizing the communication overhead. The in-depth experiment on a recent network traffic dataset shows that the proposed system achieved a high detection rate while minimizing the false positive rate, and the detection delay.
Title:
Revisiting the Performance of Deep Learning-Based Vulnerability Detection on Realistic Datasets
Abstract
The impact of software vulnerabilities on everyday software systems is significant. Despite deep learning models being proposed for vulnerability detection, their reliability is questionable. Prior evaluations show high recall/F1 scores of up to 99%, but these models underperform in practical scenarios, particularly when assessed on entire codebases rather than just the fixing commit. This paper introduces Real-Vul, a comprehensive dataset representing real-world scenarios for evaluating vulnerability detection models. Evaluating DeepWukong, LineVul, ReVeal, and IVDetect shows a significant drop in performance, with precision decreasing by up to 95 percentage points and F1 scores by up to 91 points. Furthermore, Model performance fluctuates based on vulnerability characteristics, with better F1 scores for information leaks or code injection than for path resolution or predictable return values. The results highlight a significant performance gap that needs addressing before deploying deep learning-based vulnerability detection in practical settings. Overfitting is identified as a key issue, and an augmentation technique is proposed, potentially improving performance by up to 30%. Contributions include a dataset creation approach for better model evaluation, Real-Vul dataset, and empirical evidence of deep learning models struggling in real-world settings.
Title:
A Toolchain for Comprehensive Audio/Video Analysis Using Deep Learning Based Multimodal Approach (A use case of riot or violent context detection)
Authors: Lam Pham, Phat Lam, Tin Nguyen, Hieu Tang, Alexander Schindler
Abstract
In this paper, we present a toolchain for a comprehensive audio/video analysis by leveraging deep learning based multimodal approach. To this end, different specific tasks of Speech to Text (S2T), Acoustic Scene Classification (ASC), Acoustic Event Detection (AED), Visual Object Detection (VOD), Image Captioning (IC), and Video Captioning (VC) are conducted and integrated into the toolchain. By combining individual tasks and analyzing both audio \& visual data extracted from input video, the toolchain offers various audio/video-based applications: Two general applications of audio/video clustering, comprehensive audio/video summary and a specific application of riot or violent context detection. Furthermore, the toolchain presents a flexible and adaptable architecture that is effective to integrate new models for further audio/video-based applications.
Abstract
Adversarial examples have shown a powerful ability to make a well-trained model misclassified. Current mainstream adversarial attack methods only consider one of the distortions among $L_0$-norm, $L2$-norm, and $L\infty$-norm. $L_0$-norm based methods cause large modification on a single pixel, resulting in naked-eye visible detection, while $L2$-norm and $L\infty$-norm based methods suffer from weak robustness against adversarial defense since they always diffuse tiny perturbations to all pixels. A more realistic adversarial perturbation should be sparse and imperceptible. In this paper, we propose a novel $L_p$-norm distortion-efficient adversarial attack, which not only owns the least $L_2$-norm loss but also significantly reduces the $L_0$-norm distortion. To this aim, we design a new optimization scheme, which first optimizes an initial adversarial perturbation under $L_2$-norm constraint, and then constructs a dimension unimportance matrix for the initial perturbation. Such a dimension unimportance matrix can indicate the adversarial unimportance of each dimension of the initial perturbation. Furthermore, we introduce a new concept of adversarial threshold for the dimension unimportance matrix. The dimensions of the initial perturbation whose unimportance is higher than the threshold will be all set to zero, greatly decreasing the $L_0$-norm distortion. Experimental results on three benchmark datasets show that under the same query budget, the adversarial examples generated by our method have lower $L_0$-norm and $L_2$-norm distortion than the state-of-the-art. Especially for the MNIST dataset, our attack reduces 8.1$\%$ $L_2$-norm distortion meanwhile remaining 47$\%$ pixels unattacked. This demonstrates the superiority of the proposed method over its competitors in terms of adversarial robustness and visual imperceptibility.
Title:
Towards Efficient Pixel Labeling for Industrial Anomaly Detection and Localization
Abstract
In the realm of practical Anomaly Detection (AD) tasks, manual labeling of anomalous pixels proves to be a costly endeavor. Consequently, many AD methods are crafted as one-class classifiers, tailored for training sets completely devoid of anomalies, ensuring a more cost-effective approach. While some pioneering work has demonstrated heightened AD accuracy by incorporating real anomaly samples in training, this enhancement comes at the price of labor-intensive labeling processes. This paper strikes the balance between AD accuracy and labeling expenses by introducing ADClick, a novel Interactive Image Segmentation (IIS) algorithm. ADClick efficiently generates "ground-truth" anomaly masks for real defective images, leveraging innovative residual features and meticulously crafted language prompts. Notably, ADClick showcases a significantly elevated generalization capacity compared to existing state-of-the-art IIS approaches. Functioning as an anomaly labeling tool, ADClick generates high-quality anomaly labels (AP $= 94.1\%$ on MVTec AD) based on only $3$ to $5$ manual click annotations per training image. Furthermore, we extend the capabilities of ADClick into ADClick-Seg, an enhanced model designed for anomaly detection and localization. By fine-tuning the ADClick-Seg model using the weak labels inferred by ADClick, we establish the state-of-the-art performances in supervised AD tasks (AP $= 86.4\%$ on MVTec AD and AP $= 78.4\%$, PRO $= 98.6\%$ on KSDD2).
Title:
Machine Learning Models for Improved Tracking from Range-Doppler Map Images
Authors: Elizabeth Hou, Ross Greenwood, Piyush Kumar
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Statistical tracking filters depend on accurate target measurements and uncertainty estimates for good tracking performance. In this work, we propose novel machine learning models for target detection and uncertainty estimation in range-Doppler map (RDM) images for Ground Moving Target Indicator (GMTI) radars. We show that by using the outputs of these models, we can significantly improve the performance of a multiple hypothesis tracker for complex multi-target air-to-ground tracking scenarios.
Title:
Global Context Modeling in YOLOv8 for Pediatric Wrist Fracture Detection
Abstract
Children often suffer wrist injuries in daily life, while fracture injuring radiologists usually need to analyze and interpret X-ray images before surgical treatment by surgeons. The development of deep learning has enabled neural network models to work as computer-assisted diagnosis (CAD) tools to help doctors and experts in diagnosis. Since the YOLOv8 models have obtained the satisfactory success in object detection tasks, it has been applied to fracture detection. The Global Context (GC) block effectively models the global context in a lightweight way, and incorporating it into YOLOv8 can greatly improve the model performance. This paper proposes the YOLOv8+GC model for fracture detection, which is an improved version of the YOLOv8 model with the GC block. Experimental results demonstrate that compared to the original YOLOv8 model, the proposed YOLOv8-GC model increases the mean average precision calculated at intersection over union threshold of 0.5 (mAP 50) from 63.58% to 66.32% on the GRAZPEDWRI-DX dataset, achieving the state-of-the-art (SOTA) level. The implementation code for this work is available on GitHub at this https URL.
Title:
Relating CNN-Transformer Fusion Network for Change Detection
Authors: Yuhao Gao, Gensheng Pei, Mengmeng Sheng, Zeren Sun, Tao Chen, Yazhou Yao
Abstract
While deep learning, particularly convolutional neural networks (CNNs), has revolutionized remote sensing (RS) change detection (CD), existing approaches often miss crucial features due to neglecting global context and incomplete change learning. Additionally, transformer networks struggle with low-level details. RCTNet addresses these limitations by introducing \textbf{(1)} an early fusion backbone to exploit both spatial and temporal features early on, \textbf{(2)} a Cross-Stage Aggregation (CSA) module for enhanced temporal representation, \textbf{(3)} a Multi-Scale Feature Fusion (MSF) module for enriched feature extraction in the decoder, and \textbf{(4)} an Efficient Self-deciphering Attention (ESA) module utilizing transformers to capture global information and fine-grained details for accurate change detection. Extensive experiments demonstrate RCTNet's clear superiority over traditional RS image CD methods, showing significant improvement and an optimal balance between accuracy and computational cost.
Title:
DyFADet: Dynamic Feature Aggregation for Temporal Action Detection
Authors: Le Yang, Ziwei Zheng, Yizeng Han, Hao Cheng, Shiji Song, Gao Huang, Fan Li
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Recent proposed neural network-based Temporal Action Detection (TAD) models are inherently limited to extracting the discriminative representations and modeling action instances with various lengths from complex scenes by shared-weights detection heads. Inspired by the successes in dynamic neural networks, in this paper, we build a novel dynamic feature aggregation (DFA) module that can simultaneously adapt kernel weights and receptive fields at different timestamps. Based on DFA, the proposed dynamic encoder layer aggregates the temporal features within the action time ranges and guarantees the discriminability of the extracted representations. Moreover, using DFA helps to develop a Dynamic TAD head (DyHead), which adaptively aggregates the multi-scale features with adjusted parameters and learned receptive fields better to detect the action instances with diverse ranges from videos. With the proposed encoder layer and DyHead, a new dynamic TAD model, DyFADet, achieves promising performance on a series of challenging TAD benchmarks, including HACS-Segment, THUMOS14, ActivityNet-1.3, Epic-Kitchen 100, Ego4D-Moment QueriesV1.0, and FineAction. Code is released to this https URL.
Title:
SegVG: Transferring Object Bounding Box to Segmentation for Visual Grounding
Authors: Weitai Kang, Gaowen Liu, Mubarak Shah, Yan Yan
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Different from Object Detection, Visual Grounding deals with detecting a bounding box for each text-image pair. This one box for each text-image data provides sparse supervision signals. Although previous works achieve impressive results, their passive utilization of annotation, i.e. the sole use of the box annotation as regression ground truth, results in a suboptimal performance. In this paper, we present SegVG, a novel method transfers the box-level annotation as Segmentation signals to provide an additional pixel-level supervision for Visual Grounding. Specifically, we propose the Multi-layer Multi-task Encoder-Decoder as the target grounding stage, where we learn a regression query and multiple segmentation queries to ground the target by regression and segmentation of the box in each decoding layer, respectively. This approach allows us to iteratively exploit the annotation as signals for both box-level regression and pixel-level segmentation. Moreover, as the backbones are typically initialized by pretrained parameters learned from unimodal tasks and the queries for both regression and segmentation are static learnable embeddings, a domain discrepancy remains among these three types of features, which impairs subsequent target grounding. To mitigate this discrepancy, we introduce the Triple Alignment module, where the query, text, and vision tokens are triangularly updated to share the same space by triple attention mechanism. Extensive experiments on five widely used datasets validate our state-of-the-art (SOTA) performance.
Title:
Category-Aware Dynamic Label Assignment with High-Quality Oriented Proposal
Authors: Mingkui Feng, Hancheng Yu, Xiaoyu Dang, Ming Zhou
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Objects in aerial images are typically embedded in complex backgrounds and exhibit arbitrary orientations. When employing oriented bounding boxes (OBB) to represent arbitrary oriented objects, the periodicity of angles could lead to discontinuities in label regression values at the boundaries, inducing abrupt fluctuations in the loss function. To address this problem, an OBB representation based on the complex plane is introduced in the oriented detection framework, and a trigonometric loss function is proposed. Moreover, leveraging prior knowledge of complex background environments and significant differences in large objects in aerial images, a conformer RPN head is constructed to predict angle information. The proposed loss function and conformer RPN head jointly generate high-quality oriented proposals. A category-aware dynamic label assignment based on predicted category feedback is proposed to address the limitations of solely relying on IoU for proposal label assignment. This method makes negative sample selection more representative, ensuring consistency between classification and regression features. Experiments were conducted on four realistic oriented detection datasets, and the results demonstrate superior performance in oriented object detection with minimal parameter tuning and time costs. Specifically, mean average precision (mAP) scores of 82.02%, 71.99%, 69.87%, and 98.77% were achieved on the DOTA-v1.0, DOTA-v1.5, DIOR-R, and HRSC2016 datasets, respectively.
Title:
Cyclic Refiner: Object-Aware Temporal Representation Learning for Multi-View 3D Detection and Tracking
Authors: Mingzhe Guo, Zhipeng Zhang, Liping Jing, Yuan He, Ke Wang, Heng Fan
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
We propose a unified object-aware temporal learning framework for multi-view 3D detection and tracking tasks. Having observed that the efficacy of the temporal fusion strategy in recent multi-view perception methods may be weakened by distractors and background clutters in historical frames, we propose a cyclic learning mechanism to improve the robustness of multi-view representation learning. The essence is constructing a backward bridge to propagate information from model predictions (e.g., object locations and sizes) to image and BEV features, which forms a circle with regular inference. After backward refinement, the responses of target-irrelevant regions in historical frames would be suppressed, decreasing the risk of polluting future frames and improving the object awareness ability of temporal fusion. We further tailor an object-aware association strategy for tracking based on the cyclic learning model. The cyclic learning model not only provides refined features, but also delivers finer clues (e.g., scale level) for tracklet association. The proposed cycle learning method and association module together contribute a novel and unified multi-task framework. Experiments on nuScenes show that the proposed model achieves consistent performance gains over baselines of different designs (i.e., dense query-based BEVFormer, sparse query-based SparseBEV and LSS-based BEVDet4D) on both detection and tracking evaluation.
Title:
Visual Grounding with Attention-Driven Constraint Balancing
Authors: Weitai Kang, Luowei Zhou, Junyi Wu, Changchang Sun, Yan Yan
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Unlike Object Detection, Visual Grounding task necessitates the detection of an object described by complex free-form language. To simultaneously model such complex semantic and visual representations, recent state-of-the-art studies adopt transformer-based models to fuse features from both modalities, further introducing various modules that modulate visual features to align with the language expressions and eliminate the irrelevant redundant information. However, their loss function, still adopting common Object Detection losses, solely governs the bounding box regression output, failing to fully optimize for the above objectives. To tackle this problem, in this paper, we first analyze the attention mechanisms of transformer-based models. Building upon this, we further propose a novel framework named Attention-Driven Constraint Balancing (AttBalance) to optimize the behavior of visual features within language-relevant regions. Extensive experimental results show that our method brings impressive improvements. Specifically, we achieve constant improvements over five different models evaluated on four different benchmarks. Moreover, we attain a new state-of-the-art performance by integrating our method into QRNet.
Title:
ACTRESS: Active Retraining for Semi-supervised Visual Grounding
Authors: Weitai Kang, Mengxue Qu, Yunchao Wei, Yan Yan
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Semi-Supervised Visual Grounding (SSVG) is a new challenge for its sparse labeled data with the need for multimodel understanding. A previous study, RefTeacher, makes the first attempt to tackle this task by adopting the teacher-student framework to provide pseudo confidence supervision and attention-based supervision. However, this approach is incompatible with current state-of-the-art visual grounding models, which follow the Transformer-based pipeline. These pipelines directly regress results without region proposals or foreground binary classification, rendering them unsuitable for fitting in RefTeacher due to the absence of confidence scores. Furthermore, the geometric difference in teacher and student inputs, stemming from different data augmentations, induces natural misalignment in attention-based constraints. To establish a compatible SSVG framework, our paper proposes the ACTive REtraining approach for Semi-Supervised Visual Grounding, abbreviated as ACTRESS. Initially, the model is enhanced by incorporating an additional quantized detection head to expose its detection confidence. Building upon this, ACTRESS consists of an active sampling strategy and a selective retraining strategy. The active sampling strategy iteratively selects high-quality pseudo labels by evaluating three crucial aspects: Faithfulness, Robustness, and Confidence, optimizing the utilization of unlabeled data. The selective retraining strategy retrains the model with periodic re-initialization of specific parameters, facilitating the model's escape from local minima. Extensive experiments demonstrates our superior performance on widely-used benchmark datasets.
Title:
STF: Sentence Transformer Fine-Tuning For Topic Categorization With Limited Data
Abstract
Nowadays, topic classification from tweets attracts considerable research attention. Different classification systems have been suggested thanks to these research efforts. Nevertheless, they face major challenges owing to low performance metrics due to the limited amount of labeled data. We propose Sentence Transformers Fine-tuning (STF), a topic detection system that leverages pretrained Sentence Transformers models and fine-tuning to classify topics from tweets accurately. Moreover, extensive parameter sensitivity analyses were conducted to finetune STF parameters for our topic classification task to achieve the best performance results. Experiments on two benchmark datasets demonstrated that (1) the proposed STF can be effectively used for classifying tweet topics and outperforms the latest state-of-the-art approaches, and (2) the proposed STF does not require a huge amount of labeled tweets to achieve good accuracy, which is a limitation of many state-of-the-art approaches. Our main contribution is the achievement of promising results in tweet topic classification by applying pretrained sentence transformers language models.
Title:
Anomaly-based Framework for Detecting Power Overloading Cyberattacks in Smart Grid AMI
Abstract
The Advanced Metering Infrastructure (AMI) is one of the key components of the smart grid. It provides interactive services for managing billing and electricity consumption, but it also introduces new vectors for cyberattacks. Although, the devastating and severe impact of power overloading cyberattacks on smart grid AMI, few researches in the literature have addressed them. In the present paper, we propose a two-level anomaly detection framework based on regression decision trees. The introduced detection approach leverages the regularity and predictability of energy consumption to build reference consumption patterns for the whole neighborhood and each household within it. Using a reference consumption pattern enables detecting power overloading cyberattacks regardless of the attacker's strategy as they cause a drastic change in the consumption pattern. The continuous two-level monitoring of energy consumption load allows efficient and early detection of cyberattacks. We carried out an extensive experiment on a real-world publicly available energy consumption dataset of 500 customers in Ireland. We extracted, from the raw data, the relevant attributes for training the energy consumption patterns. The evaluation shows that our approach achieves a high detection rate, a low false alarm rate, and superior performances compared to existing solutions.
Title:
BACON: Supercharge Your VLM with Bag-of-Concept Graph to Mitigate Hallucinations
Authors: Zhantao Yang, Ruili Feng, Keyu Yan, Huangji Wang, Zhicai Wang, Shangwen Zhu, Han Zhang, Jie Xiao, Pingyu Wu, Kai Zhu, Jixuan Chen, Chen-Wei Xie, Chaojie Mao, Yue Yang, Hongyang Zhang, Yu Liu, Fan Cheng
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL); Databases (cs.DB)
Abstract
This paper presents Bag-of-Concept Graph (BACON) to gift models with limited linguistic abilities to taste the privilege of Vision Language Models (VLMs) and boost downstream tasks such as detection, visual question answering (VQA), and image generation. Since the visual scenes in physical worlds are structured with complex relations between objects, BACON breaks down annotations into basic minimum elements and presents them in a graph structure. Element-wise style enables easy understanding, and structural composition liberates difficult locating. Careful prompt design births the BACON captions with the help of public-available VLMs and segmentation methods. In this way, we gather a dataset with 100K annotated images, which endow VLMs with remarkable capabilities, such as accurately generating BACON, transforming prompts into BACON format, envisioning scenarios in the style of BACONr, and dynamically modifying elements within BACON through interactive dialogue and more. Wide representative experiments, including detection, VQA, and image generation tasks, tell BACON as a lifeline to achieve previous out-of-reach tasks or excel in their current cutting-edge solutions.
Keyword: face recognition
Title:
SlerpFace: Face Template Protection via Spherical Linear Interpolation
Abstract
Contemporary face recognition systems use feature templates extracted from face images to identify persons. To enhance privacy, face template protection techniques are widely employed to conceal sensitive identity and appearance information stored in the template. This paper identifies an emerging privacy attack form utilizing diffusion models that could nullify prior protection, referred to as inversion attacks. The attack can synthesize high-quality, identity-preserving face images from templates, revealing persons' appearance. Based on studies of the diffusion model's generative capability, this paper proposes a defense to deteriorate the attack, by rotating templates to a noise-like distribution. This is achieved efficiently by spherically and linearly interpolating templates, or slerp, on their located hypersphere. This paper further proposes to group-wisely divide and drop out templates' feature dimensions, to enhance the irreversibility of rotated templates. The division of groups and dropouts within each group are learned in a recognition-favored way. The proposed techniques are concretized as a novel face template protection technique, SlerpFace. Extensive experiments show that SlerpFace provides satisfactory recognition accuracy and comprehensive privacy protection against inversion and other attack forms, superior to prior arts.
Abstract
Panoramic images, capturing a 360° field of view (FoV), encompass omnidirectional spatial information crucial for scene understanding. However, it is not only costly to obtain training-sufficient dense-annotated panoramas but also application-restricted when training models in a close-vocabulary setting. To tackle this problem, in this work, we define a new task termed Open Panoramic Segmentation (OPS), where models are trained with FoV-restricted pinhole images in the source domain in an open-vocabulary setting while evaluated with FoV-open panoramic images in the target domain, enabling the zero-shot open panoramic semantic segmentation ability of models. Moreover, we propose a model named OOOPS with a Deformable Adapter Network (DAN), which significantly improves zero-shot panoramic semantic segmentation performance. To further enhance the distortion-aware modeling ability from the pinhole source domain, we propose a novel data augmentation method called Random Equirectangular Projection (RERP) which is specifically designed to address object deformations in advance. Surpassing other state-of-the-art open-vocabulary semantic segmentation approaches, a remarkable performance boost on three panoramic datasets, WildPASS, Stanford2D3D, and Matterport3D, proves the effectiveness of our proposed OOOPS model with RERP on the OPS task, especially +2.2% on outdoor WildPASS and +2.4% mIoU on indoor Stanford2D3D. The code will be available at this https URL.
Title:
ShiftAddAug: Augment Multiplication-Free Tiny Neural Network with Hybrid Computation
Abstract
Operators devoid of multiplication, such as Shift and Add, have gained prominence for their compatibility with hardware. However, neural networks (NNs) employing these operators typically exhibit lower accuracy compared to conventional NNs with identical structures. ShiftAddAug uses costly multiplication to augment efficient but less powerful multiplication-free operators, improving performance without any inference overhead. It puts a ShiftAdd tiny NN into a large multiplicative model and encourages it to be trained as a sub-model to obtain additional supervision. In order to solve the weight discrepancy problem between hybrid operators, a new weight sharing method is proposed. Additionally, a novel two stage neural architecture search is used to obtain better augmentation effects for smaller but stronger multiplication-free tiny neural networks. The superiority of ShiftAddAug is validated through experiments in image classification and semantic segmentation, consistently delivering noteworthy enhancements. Remarkably, it secures up to a 4.95% increase in accuracy on the CIFAR100 compared to its directly trained counterparts, even surpassing the performance of multiplicative NNs.
Title:
Finding Spanning Trees with Perfect Matchings
Authors: Kristóf Bérczi, Tamás Király, Yusuke Kobayashi, Yutaro Yamaguchi, Yu Yokoi
Subjects: Subjects:
Data Structures and Algorithms (cs.DS); Combinatorics (math.CO)
Abstract
We investigate the tractability of a simple fusion of two fundamental structures on graphs, a spanning tree and a perfect matching. Specifically, we consider the following problem: given an edge-weighted graph, find a minimum-weight spanning tree among those containing a perfect matching. On the positive side, we design a simple greedy algorithm for the case when the graph is complete (or complete bipartite) and the edge weights take at most two values. On the negative side, the problem is NP-hard even when the graph is complete (or complete bipartite) and the edge weights take at most three values, or when the graph is cubic, planar, and bipartite and the edge weights take at most two values. We also consider an interesting variant. We call a tree strongly balanced if on one side of the bipartition of the vertex set with respect to the tree, all but one of the vertices have degree $2$ and the remaining one is a leaf. This property is a sufficient condition for a tree to have a perfect matching, which enjoys an additional property. When the underlying graph is bipartite, strongly balanced spanning trees can be written as matroid intersection, and this fact was recently utilized to design an approximation algorithm for some kind of connectivity augmentation problem. The natural question is its tractability in nonbipartite graphs. As a negative answer, it turns out NP-hard to test whether a given graph has a strongly balanced spanning tree or not even when the graph is subcubic and planar.
Title:
YOLOv5, YOLOv8 and YOLOv10: The Go-To Detectors for Real-time Vision
Authors: Muhammad Hussain
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
This paper presents a comprehensive review of the evolution of the YOLO (You Only Look Once) object detection algorithm, focusing on YOLOv5, YOLOv8, and YOLOv10. We analyze the architectural advancements, performance improvements, and suitability for edge deployment across these versions. YOLOv5 introduced significant innovations such as the CSPDarknet backbone and Mosaic Augmentation, balancing speed and accuracy. YOLOv8 built upon this foundation with enhanced feature extraction and anchor-free detection, improving versatility and performance. YOLOv10 represents a leap forward with NMS-free training, spatial-channel decoupled downsampling, and large-kernel convolutions, achieving state-of-the-art performance with reduced computational overhead. Our findings highlight the progressive enhancements in accuracy, efficiency, and real-time performance, particularly emphasizing their applicability in resource-constrained environments. This review provides insights into the trade-offs between model complexity and detection accuracy, offering guidance for selecting the most appropriate YOLO version for specific edge computing applications.
Title:
An Organism Starts with a Single Pix-Cell: A Neural Cellular Diffusion for High-Resolution Image Synthesis
Authors: Marawan Elbatel, Konstantinos Kamnitsas, Xiaomeng Li
Abstract
Generative modeling seeks to approximate the statistical properties of real data, enabling synthesis of new data that closely resembles the original distribution. Generative Adversarial Networks (GANs) and Denoising Diffusion Probabilistic Models (DDPMs) represent significant advancements in generative modeling, drawing inspiration from game theory and thermodynamics, respectively. Nevertheless, the exploration of generative modeling through the lens of biological evolution remains largely untapped. In this paper, we introduce a novel family of models termed Generative Cellular Automata (GeCA), inspired by the evolution of an organism from a single cell. GeCAs are evaluated as an effective augmentation tool for retinal disease classification across two imaging modalities: Fundus and Optical Coherence Tomography (OCT). In the context of OCT imaging, where data is scarce and the distribution of classes is inherently skewed, GeCA significantly boosts the performance of 11 different ophthalmological conditions, achieving a 12% increase in the average F1 score compared to conventional baselines. GeCAs outperform both diffusion methods that incorporate UNet or state-of-the art variants with transformer-based denoising models, under similar parameter constraints. Code is available at: this https URL.
Title:
ALTER: Augmentation for Large-Table-Based Reasoning
Authors: Han Zhang, Yuheng Ma, Hanfang Yang
Subjects: Subjects:
Computation and Language (cs.CL)
Abstract
While extensive research has explored the use of large language models (LLMs) for table-based reasoning, most approaches struggle with scalability when applied to large tables. To maintain the superior comprehension abilities of LLMs in these scenarios, we introduce ALTER(Augmentation for Large-Table-Based Reasoning)-a framework designed to harness the latent augmentation potential in both free-form natural language (NL) questions, via the query augmentor, and semi-structured tabular data, through the table augmentor. By utilizing only a small subset of relevant data from the table and supplementing it with pre-augmented schema, semantic, and literal information, ALTER achieves outstanding performance on table-based reasoning benchmarks. We also provide a detailed analysis of large-table scenarios, comparing different methods and various partitioning principles. In these scenarios, our method outperforms all other approaches and exhibits robustness and efficiency against perturbations.
Title:
Revisiting the Performance of Deep Learning-Based Vulnerability Detection on Realistic Datasets
Abstract
The impact of software vulnerabilities on everyday software systems is significant. Despite deep learning models being proposed for vulnerability detection, their reliability is questionable. Prior evaluations show high recall/F1 scores of up to 99%, but these models underperform in practical scenarios, particularly when assessed on entire codebases rather than just the fixing commit. This paper introduces Real-Vul, a comprehensive dataset representing real-world scenarios for evaluating vulnerability detection models. Evaluating DeepWukong, LineVul, ReVeal, and IVDetect shows a significant drop in performance, with precision decreasing by up to 95 percentage points and F1 scores by up to 91 points. Furthermore, Model performance fluctuates based on vulnerability characteristics, with better F1 scores for information leaks or code injection than for path resolution or predictable return values. The results highlight a significant performance gap that needs addressing before deploying deep learning-based vulnerability detection in practical settings. Overfitting is identified as a key issue, and an augmentation technique is proposed, potentially improving performance by up to 30%. Contributions include a dataset creation approach for better model evaluation, Real-Vul dataset, and empirical evidence of deep learning models struggling in real-world settings.
Title:
Enhancing Class Fairness in Classification with A Two-Player Game Approach
Authors: Yunpeng Jiang, Paul Weng, Yutong Ban
Subjects: Subjects:
Computers and Society (cs.CY); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Computer Science and Game Theory (cs.GT); Machine Learning (cs.LG)
Abstract
Data augmentation is widely applied and has shown its benefits in different machine learning tasks. However, as recently observed in some downstream tasks, data augmentation may introduce an unfair impact on classifications. While it can improve the performance of some classes, it can actually be detrimental for other classes, which can be problematic in some application domains. In this paper, to counteract this phenomenon, we propose a FAir Classification approach with a Two-player game (FACT). We first formulate the training of a classifier with data augmentation as a fair optimization problem, which can be further written as an adversarial two-player game. Following this formulation, we propose a novel multiplicative weight optimization algorithm, for which we theoretically prove that it can converge to a solution that is fair over classes. Interestingly, our formulation also reveals that this fairness issue over classes is not due to data augmentation only, but is in fact a general phenomenon. Our empirical experiments demonstrate that the performance of our learned classifiers is indeed more fairly distributed over classes in five datasets, with only limited impact on the average accuracy.
Title:
ACTRESS: Active Retraining for Semi-supervised Visual Grounding
Authors: Weitai Kang, Mengxue Qu, Yunchao Wei, Yan Yan
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Semi-Supervised Visual Grounding (SSVG) is a new challenge for its sparse labeled data with the need for multimodel understanding. A previous study, RefTeacher, makes the first attempt to tackle this task by adopting the teacher-student framework to provide pseudo confidence supervision and attention-based supervision. However, this approach is incompatible with current state-of-the-art visual grounding models, which follow the Transformer-based pipeline. These pipelines directly regress results without region proposals or foreground binary classification, rendering them unsuitable for fitting in RefTeacher due to the absence of confidence scores. Furthermore, the geometric difference in teacher and student inputs, stemming from different data augmentations, induces natural misalignment in attention-based constraints. To establish a compatible SSVG framework, our paper proposes the ACTive REtraining approach for Semi-Supervised Visual Grounding, abbreviated as ACTRESS. Initially, the model is enhanced by incorporating an additional quantized detection head to expose its detection confidence. Building upon this, ACTRESS consists of an active sampling strategy and a selective retraining strategy. The active sampling strategy iteratively selects high-quality pseudo labels by evaluating three crucial aspects: Faithfulness, Robustness, and Confidence, optimizing the utilization of unlabeled data. The selective retraining strategy retrains the model with periodic re-initialization of specific parameters, facilitating the model's escape from local minima. Extensive experiments demonstrates our superior performance on widely-used benchmark datasets.
Keyword: detection
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Keyword: face recognition
Title:
Keyword: augmentation
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title: