Self-supervised multimodal neuroimaging yields predictive representations for a spectrum of Alzheimer's phenotypes
Authors: Alex Fedorov, Eloy Geenjaar, Lei Wu, Tristan Sylvain, Thomas P. DeRamus, Margaux Luck, Maria Misiura, R Devon Hjelm, Sergey M. Plis, Vince D. Calhoun
Subjects: Machine Learning (cs.LG); Image and Video Processing (eess.IV); Neurons and Cognition (q-bio.NC)
Abstract
Recent neuroimaging studies that focus on predicting brain disorders via modern machine learning approaches commonly include a single modality and rely on supervised over-parameterized models.However, a single modality provides only a limited view of the highly complex brain. Critically, supervised models in clinical settings lack accurate diagnostic labels for training. Coarse labels do not capture the long-tailed spectrum of brain disorder phenotypes, which leads to a loss of generalizability of the model that makes them less useful in diagnostic settings. This work presents a novel multi-scale coordinated framework for learning multiple representations from multimodal neuroimaging data. We propose a general taxonomy of informative inductive biases to capture unique and joint information in multimodal self-supervised fusion. The taxonomy forms a family of decoder-free models with reduced computational complexity and a propensity to capture multi-scale relationships between local and global representations of the multimodal inputs. We conduct a comprehensive evaluation of the taxonomy using functional and structural magnetic resonance imaging (MRI) data across a spectrum of Alzheimer's disease phenotypes and show that self-supervised models reveal disorder-relevant brain regions and multimodal links without access to the labels during pre-training. The proposed multimodal self-supervised learning yields representations with improved classification performance for both modalities. The concomitant rich and flexible unsupervised deep learning framework captures complex multimodal relationships and provides predictive performance that meets or exceeds that of a more narrow supervised classification analysis. We present elaborate quantitative evidence of how this framework can significantly advance our search for missing links in complex brain disorders.
BiFuse++: Self-supervised and Efficient Bi-projection Fusion for 360 Depth Estimation
Authors: Fu-En Wang, Yu-Hsuan Yeh, Yi-Hsuan Tsai, Wei-Chen Chiu, Min Sun
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
Due to the rise of spherical cameras, monocular 360 depth estimation becomes an important technique for many applications (e.g., autonomous systems). Thus, state-of-the-art frameworks for monocular 360 depth estimation such as bi-projection fusion in BiFuse are proposed. To train such a framework, a large number of panoramas along with the corresponding depth ground truths captured by laser sensors are required, which highly increases the cost of data collection. Moreover, since such a data collection procedure is time-consuming, the scalability of extending these methods to different scenes becomes a challenge. To this end, self-training a network for monocular depth estimation from 360 videos is one way to alleviate this issue. However, there are no existing frameworks that incorporate bi-projection fusion into the self-training scheme, which highly limits the self-supervised performance since bi-projection fusion can leverage information from different projection types. In this paper, we propose BiFuse++ to explore the combination of bi-projection fusion and the self-training scenario. To be specific, we propose a new fusion module and Contrast-Aware Photometric Loss to improve the performance of BiFuse and increase the stability of self-training on real-world videos. We conduct both supervised and self-supervised experiments on benchmark datasets and achieve state-of-the-art performance.
MimCo: Masked Image Modeling Pre-training with Contrastive Teacher
Abstract
Recent masked image modeling (MIM) has received much attention in self-supervised learning (SSL), which requires the target model to recover the masked part of the input image. Although MIM-based pre-training methods achieve new state-of-the-art performance when transferred to many downstream tasks, the visualizations show that the learned representations are less separable, especially compared to those based on contrastive learning pre-training. This inspires us to think whether the linear separability of MIM pre-trained representation can be further improved, thereby improving the pre-training performance. Since MIM and contrastive learning tend to utilize different data augmentations and training strategies, combining these two pretext tasks is not trivial. In this work, we propose a novel and flexible pre-training framework, named MimCo, which combines MIM and contrastive learning through two-stage pre-training. Specifically, MimCo takes a pre-trained contrastive learning model as the teacher model and is pre-trained with two types of learning targets: patch-level and image-level reconstruction losses. Extensive transfer experiments on downstream tasks demonstrate the superior performance of our MimCo pre-training framework. Taking ViT-S as an example, when using the pre-trained MoCov3-ViT-S as the teacher model, MimCo only needs 100 epochs of pre-training to achieve 82.53% top-1 finetuning accuracy on Imagenet-1K, which outperforms the state-of-the-art self-supervised learning counterparts.
Abstract
Smart cars, smartphones and other devices in the Internet of Things (IoT), which usually have more than one sensors, produce multimodal data. Federated Learning supports collecting a wealth of multimodal data from different devices without sharing raw data. Transfer Learning methods help transfer knowledge from some devices to others. Federated Transfer Learning methods benefit both Federated Learning and Transfer Learning. This newly proposed Federated Transfer Learning framework aims at connecting data islands with privacy protection. Our construction is based on Federated Learning and Transfer Learning. Compared with previous Federated Transfer Learnings, where each user should have data with identical modalities (either all unimodal or all multimodal), our new framework is more generic, it allows a hybrid distribution of user data. The core strategy is to use two different but inherently connected training methods for our two types of users. Supervised Learning is adopted for users with only unimodal data (Type 1), while Self-Supervised Learning is applied to user with multimodal data (Type 2) for both the feature of each modality and the connection between them. This connection knowledge of Type 2 will help Type 1 in later stages of training. Training in the new framework can be divided in three steps. In the first step, users who have data with the identical modalities are grouped together. For example, user with only sound signals are in group one, and those with only images are in group two, and users with multimodal data are in group three, and so on. In the second step, Federated Learning is executed within the groups, where Supervised Learning and Self-Supervised Learning are used depending on the group's nature. Most of the Transfer Learning happens in the third step, where the related parts in the network obtained from the previous steps are aggregated (federated).
Network Intrusion Detection with Limited Labeled Data
Authors: S. Lotfi, M. Modirrousta, S. Shashaani, S. Amini, M. Aliyari Shoorehdeli
Abstract
With the increasing dependency of daily life over computer networks, the importance of these networks security becomes prominent. Different intrusion attacks to networks have been designed and the attackers are working on improving them. Thus the ability to detect intrusion with limited number of labeled data is desirable to provide networks with higher level of security. In this paper we design an intrusion detection system based on a deep neural network. The proposed system is based on self-supervised contrastive learning where a huge amount of unlabeled data can be used to generate informative representation suitable for various downstream tasks with limited number of labeled data. Using different experiments, we have shown that the proposed system presents an accuracy of 94.05% over the UNSW-NB15 dataset, an improvement of 4.22% in comparison to previous method based on self-supervised learning. Our simulations have also shown impressive results when the size of labeled training data is limited. The performance of the resulting Encoder Block trained on UNSW-NB15 dataset has also been tested on other datasets for representation extraction which shows competitive results in downstream tasks.
Measuring the Interpretability of Unsupervised Representations via Quantized Reverse Probing
Authors: Iro Laina, Yuki M. Asano, Andrea Vedaldi
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
Self-supervised visual representation learning has recently attracted significant research interest. While a common way to evaluate self-supervised representations is through transfer to various downstream tasks, we instead investigate the problem of measuring their interpretability, i.e. understanding the semantics encoded in raw representations. We formulate the latter as estimating the mutual information between the representation and a space of manually labelled concepts. To quantify this we introduce a decoding bottleneck: information must be captured by simple predictors, mapping concepts to clusters in representation space. This approach, which we call reverse linear probing, provides a single number sensitive to the semanticity of the representation. This measure is also able to detect when the representation contains combinations of concepts (e.g., "red apple") instead of just individual attributes ("red" and "apple" independently). Finally, we propose to use supervised classifiers to automatically label large datasets in order to enrich the space of concepts used for probing. We use our method to evaluate a large number of self-supervised representations, ranking them by interpretability, highlight the differences that emerge compared to the standard evaluation with linear probes and discuss several qualitative insights. Code at: {\scriptsize{\url{https://github.com/iro-cp/ssl-qrp}}}.
Keyword: vision transformer
Fusion of Satellite Images and Weather Data with Transformer Networks for Downy Mildew Disease Detection
Authors: William Maillet, Maryam Ouhami, Adel Hafiane
Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
Abstract
Crop diseases significantly affect the quantity and quality of agricultural production. In a context where the goal of precision agriculture is to minimize or even avoid the use of pesticides, weather and remote sensing data with deep learning can play a pivotal role in detecting crop diseases, allowing localized treatment of crops. However, combining heterogeneous data such as weather and images remains a hot topic and challenging task. Recent developments in transformer architectures have shown the possibility of fusion of data from different domains, for instance text-image. The current trend is to custom only one transformer to create a multimodal fusion model. Conversely, we propose a new approach to realize data fusion using three transformers. In this paper, we first solved the missing satellite images problem, by interpolating them with a ConvLSTM model. Then, proposed a multimodal fusion architecture that jointly learns to process visual and weather information. The architecture is built from three main components, a Vision Transformer and two transformer-encoders, allowing to fuse both image and weather modalities. The results of the proposed method are promising achieving 97\% overall accuracy.
Keyword: multimodal
Fusion of Satellite Images and Weather Data with Transformer Networks for Downy Mildew Disease Detection
Authors: William Maillet, Maryam Ouhami, Adel Hafiane
Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
Abstract
Crop diseases significantly affect the quantity and quality of agricultural production. In a context where the goal of precision agriculture is to minimize or even avoid the use of pesticides, weather and remote sensing data with deep learning can play a pivotal role in detecting crop diseases, allowing localized treatment of crops. However, combining heterogeneous data such as weather and images remains a hot topic and challenging task. Recent developments in transformer architectures have shown the possibility of fusion of data from different domains, for instance text-image. The current trend is to custom only one transformer to create a multimodal fusion model. Conversely, we propose a new approach to realize data fusion using three transformers. In this paper, we first solved the missing satellite images problem, by interpolating them with a ConvLSTM model. Then, proposed a multimodal fusion architecture that jointly learns to process visual and weather information. The architecture is built from three main components, a Vision Transformer and two transformer-encoders, allowing to fuse both image and weather modalities. The results of the proposed method are promising achieving 97\% overall accuracy.
Self-supervised multimodal neuroimaging yields predictive representations for a spectrum of Alzheimer's phenotypes
Authors: Alex Fedorov, Eloy Geenjaar, Lei Wu, Tristan Sylvain, Thomas P. DeRamus, Margaux Luck, Maria Misiura, R Devon Hjelm, Sergey M. Plis, Vince D. Calhoun
Subjects: Machine Learning (cs.LG); Image and Video Processing (eess.IV); Neurons and Cognition (q-bio.NC)
Abstract
Recent neuroimaging studies that focus on predicting brain disorders via modern machine learning approaches commonly include a single modality and rely on supervised over-parameterized models.However, a single modality provides only a limited view of the highly complex brain. Critically, supervised models in clinical settings lack accurate diagnostic labels for training. Coarse labels do not capture the long-tailed spectrum of brain disorder phenotypes, which leads to a loss of generalizability of the model that makes them less useful in diagnostic settings. This work presents a novel multi-scale coordinated framework for learning multiple representations from multimodal neuroimaging data. We propose a general taxonomy of informative inductive biases to capture unique and joint information in multimodal self-supervised fusion. The taxonomy forms a family of decoder-free models with reduced computational complexity and a propensity to capture multi-scale relationships between local and global representations of the multimodal inputs. We conduct a comprehensive evaluation of the taxonomy using functional and structural magnetic resonance imaging (MRI) data across a spectrum of Alzheimer's disease phenotypes and show that self-supervised models reveal disorder-relevant brain regions and multimodal links without access to the labels during pre-training. The proposed multimodal self-supervised learning yields representations with improved classification performance for both modalities. The concomitant rich and flexible unsupervised deep learning framework captures complex multimodal relationships and provides predictive performance that meets or exceeds that of a more narrow supervised classification analysis. We present elaborate quantitative evidence of how this framework can significantly advance our search for missing links in complex brain disorders.
VGStore: A Multimodal Extension to SPARQL for Querying RDF Scene Graph
Authors: Yanzeng Li, Zilong Zheng, Wenjuan Han, Lei Zou
Subjects: Databases (cs.DB); Computation and Language (cs.CL)
Abstract
Semantic Web technology has successfully facilitated many RDF models with rich data representation methods. It also has the potential ability to represent and store multimodal knowledge bases such as multimodal scene graphs. However, most existing query languages, especially SPARQL, barely explore the implicit multimodal relationships like semantic similarity, spatial relations, etc. We first explored this issue by organizing a large-scale scene graph dataset, namely Visual Genome, in the RDF graph database. Based on the proposed RDF-stored multimodal scene graph, we extended SPARQL queries to answer questions containing relational reasoning about color, spatial, etc. Further demo (i.e., VGStore) shows the effectiveness of customized queries and displaying multimodal data.
Improving the Cross-Lingual Generalisation in Visual Question Answering
Abstract
While several benefits were realized for multilingual vision-language pretrained models, recent benchmarks across various tasks and languages showed poor cross-lingual generalisation when multilingually pre-trained vision-language models are applied to non-English data, with a large gap between (supervised) English performance and (zero-shot) cross-lingual transfer. In this work, we explore the poor performance of these models on a zero-shot cross-lingual visual question answering (VQA) task, where models are fine-tuned on English visual-question data and evaluated on 7 typologically diverse languages. We improve cross-lingual transfer with three strategies: (1) we introduce a linguistic prior objective to augment the cross-entropy loss with a similarity-based loss to guide the model during training, (2) we learn a task-specific subnetwork that improves cross-lingual generalisation and reduces variance without model modification, (3) we augment training examples using synthetic code-mixing to promote alignment of embeddings between source and target languages. Our experiments on xGQA using the pretrained multilingual multimodal transformers UC2 and M3P demonstrate the consistent effectiveness of the proposed fine-tuning strategy for 7 languages, outperforming existing transfer methods with sparse models. Code and data to reproduce our findings are publicly available.
Benchmarking Multimodal Variational Autoencoders: GeBiD Dataset and Toolkit
Authors: Gabriela Sejnova, Michal Vavrecka, Karla Stepanova
Abstract
Multimodal Variational Autoencoders (VAEs) have been a subject of intense research in the past years as they can integrate multiple modalities into a joint representation and can thus serve as a promising tool for both data classification and generation. Several approaches toward multimodal VAE learning have been proposed so far, their comparison and evaluation have however been rather inconsistent. One reason is that the models differ at the implementation level, another problem is that the datasets commonly used in these cases were not initially designed for the evaluation of multimodal generative models. This paper addresses both mentioned issues. First, we propose a toolkit for systematic multimodal VAE training and comparison. Second, we present a synthetic bimodal dataset designed for a comprehensive evaluation of the joint generation and cross-generation capabilities. We demonstrate the utility of the dataset by comparing state-of-the-art models.
DM$^2$S$^2$: Deep Multi-Modal Sequence Sets with Hierarchical Modality Attention
Abstract
There is increasing interest in the use of multimodal data in various web applications, such as digital advertising and e-commerce. Typical methods for extracting important information from multimodal data rely on a mid-fusion architecture that combines the feature representations from multiple encoders. However, as the number of modalities increases, several potential problems with the mid-fusion model structure arise, such as an increase in the dimensionality of the concatenated multimodal features and missing modalities. To address these problems, we propose a new concept that considers multimodal inputs as a set of sequences, namely, deep multimodal sequence sets (DM$^2$S$^2$). Our set-aware concept consists of three components that capture the relationships among multiple modalities: (a) a BERT-based encoder to handle the inter- and intra-order of elements in the sequences, (b) intra-modality residual attention (IntraMRA) to capture the importance of the elements in a modality, and (c) inter-modality residual attention (InterMRA) to enhance the importance of elements with modality-level granularity further. Our concept exhibits performance that is comparable to or better than the previous set-aware models. Furthermore, we demonstrate that the visualization of the learned InterMRA and IntraMRA weights can provide an interpretation of the prediction results.
Abstract
Smart cars, smartphones and other devices in the Internet of Things (IoT), which usually have more than one sensors, produce multimodal data. Federated Learning supports collecting a wealth of multimodal data from different devices without sharing raw data. Transfer Learning methods help transfer knowledge from some devices to others. Federated Transfer Learning methods benefit both Federated Learning and Transfer Learning. This newly proposed Federated Transfer Learning framework aims at connecting data islands with privacy protection. Our construction is based on Federated Learning and Transfer Learning. Compared with previous Federated Transfer Learnings, where each user should have data with identical modalities (either all unimodal or all multimodal), our new framework is more generic, it allows a hybrid distribution of user data. The core strategy is to use two different but inherently connected training methods for our two types of users. Supervised Learning is adopted for users with only unimodal data (Type 1), while Self-Supervised Learning is applied to user with multimodal data (Type 2) for both the feature of each modality and the connection between them. This connection knowledge of Type 2 will help Type 1 in later stages of training. Training in the new framework can be divided in three steps. In the first step, users who have data with the identical modalities are grouped together. For example, user with only sound signals are in group one, and those with only images are in group two, and users with multimodal data are in group three, and so on. In the second step, Federated Learning is executed within the groups, where Supervised Learning and Self-Supervised Learning are used depending on the group's nature. Most of the Transfer Learning happens in the third step, where the related parts in the network obtained from the previous steps are aggregated (federated).
Multimodal Speech Enhancement Using Burst Propagation
Authors: Leandro A. Passos, Ahmed Khubaib, Mohsin Raza, Ahsan Adeel
Abstract
This paper proposes the MBURST, a novel multimodal solution for audio-visual speech enhancements that consider the most recent neurological discoveries regarding pyramidal cells of the prefrontal cortex and other brain regions. The so-called burst propagation implements several criteria to address the credit assignment problem in a more biologically plausible manner: steering the sign and magnitude of plasticity through feedback, multiplexing the feedback and feedforward information across layers through different weight connections, approximating feedback and feedforward connections, and linearizing the feedback signals. MBURST benefits from such capabilities to learn correlations between the noisy signal and the visual stimuli, thus attributing meaning to the speech by amplifying relevant information and suppressing noise. Experiments conducted over a Grid Corpus and CHiME3-based dataset show that MBURST can reproduce similar mask reconstructions to the multimodal backpropagation-based baseline while demonstrating outstanding energy efficiency management, reducing the neuron firing rates to values up to \textbf{$70\%$} lower. Such a feature implies more sustainable implementations, suitable and desirable for hearing aids or any other similar embedded systems.
Geometric multimodal representation learning
Authors: Yasha Ektefaie, George Dasoulas, Ayush Noori, Maha Farhat, Marinka Zitnik
Abstract
Graph-centric artificial intelligence (graph AI) has achieved remarkable success in modeling interacting systems prevalent in nature, from dynamical systems in biology to particle physics. The increasing heterogeneity of data calls for graph neural architectures that can combine multiple inductive biases. However, combining data from various sources is challenging because appropriate inductive bias may vary by data modality. Multimodal learning methods fuse multiple data modalities while leveraging cross-modal dependencies to address this challenge. Here, we survey 140 studies in graph-centric AI and realize that diverse data types are increasingly brought together using graphs and fed into sophisticated multimodal models. These models stratify into image-, language-, and knowledge-grounded multimodal learning. We put forward an algorithmic blueprint for multimodal graph learning based on this categorization. The blueprint serves as a way to group state-of-the-art architectures that treat multimodal data by choosing appropriately four different components. This effort can pave the way for standardizing the design of sophisticated multimodal architectures for highly complex real-world problems.
Keyword: CLIP
AI Illustrator: Translating Raw Descriptions into Images by Prompt-based Cross-Modal Generation
Authors: Yiyang Ma, Huan Yang, Bei Liu, Jianlong Fu, Jiaying Liu
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
AI illustrator aims to automatically design visually appealing images for books to provoke rich thoughts and emotions. To achieve this goal, we propose a framework for translating raw descriptions with complex semantics into semantically corresponding images. The main challenge lies in the complexity of the semantics of raw descriptions, which may be hard to be visualized (\textit{e}.\textit{g}., "gloomy" or "Asian"). It usually poses challenges for existing methods to handle such descriptions. To address this issue, we propose a \textbf{P}rompt-based \textbf{C}ross-\textbf{M}odal Generation \textbf{Frame}work (PCM-Frame) to leverage two powerful pre-trained models, including CLIP and StyleGAN. Our framework consists of two components: a projection module from \textit{Text Embedding}s to \textit{Image Embedding}s based on prompts, and an adapted image generation module built on StyleGAN which takes \textit{Image Embedding}s as inputs and is trained by combined semantic consistency losses. To bridge the gap between realistic images and illustration designs, we further adopt a stylization model as post-processing in our framework for better visual effects. Benefiting from the pre-trained models, our method can handle complex descriptions and does not require external paired data for training. Furthermore, we have built a benchmark that consists of 200 raw descriptions. We conduct a user study to demonstrate our superiority over the competing methods with complicated texts. We release our code at https://github.com/researchmm/AI\_Illustrator}{https://github.com/researchmm/AI\_Illustrator
Measurement of the Usage of Web Clips in Underground Economy
Abstract
In this paper, we study the ecosystem of the abused Web Clips in underground economy. Through this study, we find the Web Clips is wildly used by perpetrators to penetrate iOS devices to gain profit. This work starts with 1,800 user complaint documents about cyber crimes over Web Clips. We firstly look into the ecosystem of abused Web Clips and point out the main participants and workflow. In addition, what is the Web Clips used for is demystified. Then the main participants, including creators, distributors, and operators are deeply studied based on our dataset. We try to reveal the prominent features of the illicit Web Clips and give some mitigation measures. Analysis reveals that 1) SSL certificate is overwhelmingly preferred for signing Web Clips instances compared with certificate issued by Apple. The wildly used SSL certificates can be aggregated into a limited group. 2) The content of the abused Web Clips falls into a few categories, Gambling',Fraud', and `Pornography' are among the top categories. 3) Instant messenger (IM) and live streaming platform are the most popular medium to trick victims into deploying the Web Clips. 4) The Web Clips are operated by a small amount of perpetrators, and the perpetrators tend to evade detection by taking technical approach, such as registering domain names through oversea domain name service provider, preferring easy-to-acquire new gTLD (global Top Level Domain), and deploying anti-crawler tricks. Our study gives hints on investigation of cyber crime over Web Clips, we hope that this work can help stakeholders to stay ahead of the threat.
Keyword: metric learning
There is no result
Keyword: image retrieval
There is no result
Keyword: self-supervised
Self-supervised multimodal neuroimaging yields predictive representations for a spectrum of Alzheimer's phenotypes
BiFuse++: Self-supervised and Efficient Bi-projection Fusion for 360 Depth Estimation
MimCo: Masked Image Modeling Pre-training with Contrastive Teacher
Federated Transfer Learning with Multimodal Data
Network Intrusion Detection with Limited Labeled Data
Measuring the Interpretability of Unsupervised Representations via Quantized Reverse Probing
Keyword: vision transformer
Fusion of Satellite Images and Weather Data with Transformer Networks for Downy Mildew Disease Detection
Keyword: multimodal
Fusion of Satellite Images and Weather Data with Transformer Networks for Downy Mildew Disease Detection
Self-supervised multimodal neuroimaging yields predictive representations for a spectrum of Alzheimer's phenotypes
VGStore: A Multimodal Extension to SPARQL for Querying RDF Scene Graph
Improving the Cross-Lingual Generalisation in Visual Question Answering
Benchmarking Multimodal Variational Autoencoders: GeBiD Dataset and Toolkit
DM$^2$S$^2$: Deep Multi-Modal Sequence Sets with Hierarchical Modality Attention
Federated Transfer Learning with Multimodal Data
Multimodal Speech Enhancement Using Burst Propagation
Geometric multimodal representation learning
Keyword: CLIP
AI Illustrator: Translating Raw Descriptions into Images by Prompt-based Cross-Modal Generation
Measurement of the Usage of Web Clips in Underground Economy
Gambling',Fraud', and `Pornography' are among the top categories. 3) Instant messenger (IM) and live streaming platform are the most popular medium to trick victims into deploying the Web Clips. 4) The Web Clips are operated by a small amount of perpetrators, and the perpetrators tend to evade detection by taking technical approach, such as registering domain names through oversea domain name service provider, preferring easy-to-acquire new gTLD (global Top Level Domain), and deploying anti-crawler tricks. Our study gives hints on investigation of cyber crime over Web Clips, we hope that this work can help stakeholders to stay ahead of the threat.Keyword: DALLE
There is no result