Abstract
Computing the gradients of a rendering process is paramount for diverse applications in computer vision and graphics. However, accurate computation of these gradients is challenging due to discontinuities and rendering approximations, particularly for surface-based representations and rasterization-based rendering. We present a novel method for computing gradients at visibility discontinuities for rasterization-based differentiable renderers. Our method elegantly simplifies the traditionally complex problem through a carefully designed approximation strategy, allowing for a straightforward, effective, and performant solution. We introduce a novel concept of micro-edges, which allows us to treat the rasterized images as outcomes of a differentiable, continuous process aligned with the inherently non-differentiable, discrete-pixel rasterization. This technique eliminates the necessity for rendering approximations or other modifications to the forward pass, preserving the integrity of the rendered image, which makes it applicable to rasterized masks, depth, and normals images where filtering is prohibitive. Utilizing micro-edges simplifies gradient interpretation at discontinuities and enables handling of geometry intersections, offering an advantage over the prior art. We showcase our method in dynamic human head scene reconstruction, demonstrating effective handling of camera images and segmentation masks.
Title:
Few-Shot Fruit Segmentation via Transfer Learning
Authors: Jordan A. James, Heather K. Manching, Amanda M. Hulse-Kemp, William J. Beksi
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
Abstract
Advancements in machine learning, computer vision, and robotics have paved the way for transformative solutions in various domains, particularly in agriculture. For example, accurate identification and segmentation of fruits from field images plays a crucial role in automating jobs such as harvesting, disease detection, and yield estimation. However, achieving robust and precise infield fruit segmentation remains a challenging task since large amounts of labeled data are required to handle variations in fruit size, shape, color, and occlusion. In this paper, we develop a few-shot semantic segmentation framework for infield fruits using transfer learning. Concretely, our work is aimed at addressing agricultural domains that lack publicly available labeled data. Motivated by similar success in urban scene parsing, we propose specialized pre-training using a public benchmark dataset for fruit transfer learning. By leveraging pre-trained neural networks, accurate semantic segmentation of fruit in the field is achieved with only a few labeled images. Furthermore, we show that models with pre-training learn to distinguish between fruit still on the trees and fruit that have fallen on the ground, and they can effectively transfer the knowledge to the target fruit dataset.
Title:
Boosting 3D Neuron Segmentation with 2D Vision Transformer Pre-trained on Natural Images
Authors: Yik San Cheng, Runkai Zhao, Heng Wang, Hanchuan Peng, Weidong Cai
Abstract
Neuron reconstruction, one of the fundamental tasks in neuroscience, rebuilds neuronal morphology from 3D light microscope imaging data. It plays a critical role in analyzing the structure-function relationship of neurons in the nervous system. However, due to the scarcity of neuron datasets and high-quality SWC annotations, it is still challenging to develop robust segmentation methods for single neuron reconstruction. To address this limitation, we aim to distill the consensus knowledge from massive natural image data to aid the segmentation model in learning the complex neuron structures. Specifically, in this work, we propose a novel training paradigm that leverages a 2D Vision Transformer model pre-trained on large-scale natural images to initialize our Transformer-based 3D neuron segmentation model with a tailored 2D-to-3D weight transferring strategy. Our method builds a knowledge sharing connection between the abundant natural and the scarce neuron image domains to improve the 3D neuron segmentation ability in a data-efficiency manner. Evaluated on a popular benchmark, BigNeuron, our method enhances neuron segmentation performance by 8.71% over the model trained from scratch with the same amount of training samples.
Title:
MMEarth: Exploring Multi-Modal Pretext Tasks For Geospatial Representation Learning
Authors: Vishal Nedungadi, Ankit Kariryaa, Stefan Oehmcke, Serge Belongie, Christian Igel, Nico Lang
Abstract
The volume of unlabelled Earth observation (EO) data is huge, but many important applications lack labelled training data. However, EO data offers the unique opportunity to pair data from different modalities and sensors automatically based on geographic location and time, at virtually no human labor cost. We seize this opportunity to create a diverse multi-modal pretraining dataset at global scale. Using this new corpus of 1.2 million locations, we propose a Multi-Pretext Masked Autoencoder (MP-MAE) approach to learn general-purpose representations for optical satellite images. Our approach builds on the ConvNeXt V2 architecture, a fully convolutional masked autoencoder (MAE). Drawing upon a suite of multi-modal pretext tasks, we demonstrate that our MP-MAE approach outperforms both MAEs pretrained on ImageNet and MAEs pretrained on domain-specific satellite images. This is shown on several downstream tasks including image classification and semantic segmentation. We find that multi-modal pretraining notably improves the linear probing performance, e.g. 4pp on BigEarthNet and 16pp on So2Sat, compared to pretraining on optical satellite images only. We show that this also leads to better label and parameter efficiency which are crucial aspects in global scale applications.
Title:
Adaptive Guidance Learning for Camouflaged Object Detection
Authors: Zhennan Chen, Xuying Zhang, Tian-Zhu Xiang, Ying Tai
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Camouflaged object detection (COD) aims to segment objects visually embedded in their surroundings, which is a very challenging task due to the high similarity between the objects and the background. To address it, most methods often incorporate additional information (e.g., boundary, texture, and frequency clues) to guide feature learning for better detecting camouflaged objects from the background. Although progress has been made, these methods are basically individually tailored to specific auxiliary cues, thus lacking adaptability and not consistently achieving high segmentation performance. To this end, this paper proposes an adaptive guidance learning network, dubbed \textit{AGLNet}, which is a unified end-to-end learnable model for exploring and adapting different additional cues in CNN models to guide accurate camouflaged feature learning. Specifically, we first design a straightforward additional information generation (AIG) module to learn additional camouflaged object cues, which can be adapted for the exploration of effective camouflaged features. Then we present a hierarchical feature combination (HFC) module to deeply integrate additional cues and image features to guide camouflaged feature learning in a multi-level fusion manner.Followed by a recalibration decoder (RD), different features are further aggregated and refined for accurate object prediction. Extensive experiments on three widely used COD benchmark datasets demonstrate that the proposed method achieves significant performance improvements under different additional cues, and outperforms the recent 20 state-of-the-art methods by a large margin. Our code will be made publicly available at: \textcolor{blue}{this https URL}.
Title:
SalFAU-Net: Saliency Fusion Attention U-Net for Salient Object Detection
Authors: Kassaw Abraham Mulat, Zhengyong Feng, Tegegne Solomon Eshetie, Ahmed Endris Hasen
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Salient object detection (SOD) remains an important task in computer vision, with applications ranging from image segmentation to autonomous driving. Fully convolutional network (FCN)-based methods have made remarkable progress in visual saliency detection over the last few decades. However, these methods have limitations in accurately detecting salient objects, particularly in challenging scenes with multiple objects, small objects, or objects with low resolutions. To address this issue, we proposed a Saliency Fusion Attention U-Net (SalFAU-Net) model, which incorporates a saliency fusion module into each decoder block of the attention U-net model to generate saliency probability maps from each decoder block. SalFAU-Net employs an attention mechanism to selectively focus on the most informative regions of an image and suppress non-salient regions. We train SalFAU-Net on the DUTS dataset using a binary cross-entropy loss function. We conducted experiments on six popular SOD evaluation datasets to evaluate the effectiveness of the proposed method. The experimental results demonstrate that our method, SalFAU-Net, achieves competitive performance compared to other methods in terms of mean absolute error (MAE), F-measure, s-measure, and e-measure.
Title:
AC-MAMBASEG: An adaptive convolution and Mamba-based architecture for enhanced skin lesion segmentation
Abstract
Skin lesion segmentation is a critical task in computer-aided diagnosis systems for dermatological diseases. Accurate segmentation of skin lesions from medical images is essential for early detection, diagnosis, and treatment planning. In this paper, we propose a new model for skin lesion segmentation namely AC-MambaSeg, an enhanced model that has the hybrid CNN-Mamba backbone, and integrates advanced components such as Convolutional Block Attention Module (CBAM), Attention Gate, and Selective Kernel Bottleneck. AC-MambaSeg leverages the Vision Mamba framework for efficient feature extraction, while CBAM and Selective Kernel Bottleneck enhance its ability to focus on informative regions and suppress background noise. We evaluate the performance of AC-MambaSeg on diverse datasets of skin lesion images including ISIC-2018 and PH2; then compare it against existing segmentation methods. Our model shows promising potential for improving computer-aided diagnosis systems and facilitating early detection and treatment of dermatological diseases. Our source code will be made available at: this https URL.
Title:
PTQ4SAM: Post-Training Quantization for Segment Anything
Authors: Chengtao Lv, Hong Chen, Jinyang Guo, Yifu Ding, Xianglong Liu
Abstract
Segment Anything Model (SAM) has achieved impressive performance in many computer vision tasks. However, as a large-scale model, the immense memory and computation costs hinder its practical deployment. In this paper, we propose a post-training quantization (PTQ) framework for Segment Anything Model, namely PTQ4SAM. First, we investigate the inherent bottleneck of SAM quantization attributed to the bimodal distribution in post-Key-Linear activations. We analyze its characteristics from both per-tensor and per-channel perspectives, and propose a Bimodal Integration strategy, which utilizes a mathematically equivalent sign operation to transform the bimodal distribution into a relatively easy-quantized normal distribution offline. Second, SAM encompasses diverse attention mechanisms (i.e., self-attention and two-way cross-attention), resulting in substantial variations in the post-Softmax distributions. Therefore, we introduce an Adaptive Granularity Quantization for Softmax through searching the optimal power-of-two base, which is hardware-friendly. Extensive experimental results across various vision tasks (instance segmentation, semantic segmentation and object detection), datasets and model variants show the superiority of PTQ4SAM. For example, when quantizing SAM-L to 6-bit, we achieve lossless accuracy for instance segmentation, about 0.5\% drop with theoretical 3.9$\times$ acceleration. The code is available at \url{this https URL}.
Title:
StyleSeg V2: Towards Robust One-shot Segmentation of Brain Tissue via Optimization-free Registration Error Perception
Abstract
One-shot segmentation of brain tissue requires training registration-segmentation (reg-seg) dual-model iteratively, where reg-model aims to provide pseudo masks of unlabeled images for seg-model by warping a carefully-labeled atlas. However, the imperfect reg-model induces image-mask misalignment, poisoning the seg-model subsequently. Recent StyleSeg bypasses this bottleneck by replacing the unlabeled images with their warped copies of atlas, but needs to borrow the diverse image patterns via style transformation. Here, we present StyleSeg V2, inherited from StyleSeg but granted the ability of perceiving the registration errors. The motivation is that good registration behaves in a mirrored fashion for mirrored images. Therefore, almost at no cost, StyleSeg V2 can have reg-model itself "speak out" incorrectly-aligned regions by simply mirroring (symmetrically flipping the brain) its input, and the registration errors are symmetric inconsistencies between the outputs of original and mirrored inputs. Consequently, StyleSeg V2 allows the seg-model to make use of correctly-aligned regions of unlabeled images and also enhances the fidelity of style-transformed warped atlas image by weighting the local transformation strength according to registration errors. The experimental results on three public datasets demonstrate that our proposed StyleSeg V2 outperforms other state-of-the-arts by considerable margins, and exceeds StyleSeg by increasing the average Dice by at least 2.4%.
Title:
On the Theory of Cross-Modality Distillation with Contrastive Learning
Abstract
Cross-modality distillation arises as an important topic for data modalities containing limited knowledge such as depth maps and high-quality sketches. Such techniques are of great importance, especially for memory and privacy-restricted scenarios where labeled training data is generally unavailable. To solve the problem, existing label-free methods leverage a few pairwise unlabeled data to distill the knowledge by aligning features or statistics between the source and target modalities. For instance, one typically aims to minimize the L2 distance or contrastive loss between the learned features of pairs of samples in the source (e.g. image) and the target (e.g. sketch) modalities. However, most algorithms in this domain only focus on the experimental results but lack theoretical insight. To bridge the gap between the theory and practical method of cross-modality distillation, we first formulate a general framework of cross-modality contrastive distillation (CMCD), built upon contrastive learning that leverages both positive and negative correspondence, towards a better distillation of generalizable features. Furthermore, we establish a thorough convergence analysis that reveals that the distance between source and target modalities significantly impacts the test error on downstream tasks within the target modality which is also validated by the empirical results. Extensive experimental results show that our algorithm outperforms existing algorithms consistently by a margin of 2-3\% across diverse modalities and tasks, covering modalities of image, sketch, depth map, and audio and tasks of recognition and segmentation.
Title:
3D LiDAR Mapping in Dynamic Environments Using a 4D Implicit Neural Representation
Abstract
Building accurate maps is a key building block to enable reliable localization, planning, and navigation of autonomous vehicles. We propose a novel approach for building accurate maps of dynamic environments utilizing a sequence of LiDAR scans. To this end, we propose encoding the 4D scene into a novel spatio-temporal implicit neural map representation by fitting a time-dependent truncated signed distance function to each point. Using our representation, we extract the static map by filtering the dynamic parts. Our neural representation is based on sparse feature grids, a globally shared decoder, and time-dependent basis functions, which we jointly optimize in an unsupervised fashion. To learn this representation from a sequence of LiDAR scans, we design a simple yet efficient loss function to supervise the map optimization in a piecewise way. We evaluate our approach on various scenes containing moving objects in terms of the reconstruction quality of static maps and the segmentation of dynamic point clouds. The experimental results demonstrate that our method is capable of removing the dynamic part of the input point clouds while reconstructing accurate and complete 3D maps, outperforming several state-of-the-art methods. Codes are available at: this https URL
Title:
Implantable Adaptive Cells: differentiable architecture search to improve the performance of any trained U-shaped network
Authors: Emil Benedykciuk, Marcin Denkowski, Grzegorz Wójcik
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
This paper introduces a novel approach to enhance the performance of pre-trained neural networks in medical image segmentation using Neural Architecture Search (NAS) methods, specifically Differentiable Architecture Search (DARTS). We present the concept of Implantable Adaptive Cell (IAC), small but powerful modules identified through Partially-Connected DARTS, designed to be injected into the skip connections of an existing and already trained U-shaped model. Our strategy allows for the seamless integration of the IAC into the pre-existing architecture, thereby enhancing its performance without necessitating a complete retraining from scratch. The empirical studies, focusing on medical image segmentation tasks, demonstrate the efficacy of this method. The integration of specialized IAC cells into various configurations of the U-Net model increases segmentation accuracy by almost 2\% points on average for the validation dataset and over 3\% points for the training dataset. The findings of this study not only offer a cost-effective alternative to the complete overhaul of complex models for performance upgrades but also indicate the potential applicability of our method to other architectures and problem domains.
Title:
DBDH: A Dual-Branch Dual-Head Neural Network for Invisible Embedded Regions Localization
Abstract
Embedding invisible hyperlinks or hidden codes in images to replace QR codes has become a hot topic recently. This technology requires first localizing the embedded region in the captured photos before decoding. Existing methods that train models to find the invisible embedded region struggle to obtain accurate localization results, leading to degraded decoding accuracy. This limitation is primarily because the CNN network is sensitive to low-frequency signals, while the embedded signal is typically in the high-frequency form. Based on this, this paper proposes a Dual-Branch Dual-Head (DBDH) neural network tailored for the precise localization of invisible embedded regions. Specifically, DBDH uses a low-level texture branch containing 62 high-pass filters to capture the high-frequency signals induced by embedding. A high-level context branch is used to extract discriminative features between the embedded and normal regions. DBDH employs a detection head to directly detect the four vertices of the embedding region. In addition, we introduce an extra segmentation head to segment the mask of the embedding region during training. The segmentation head provides pixel-level supervision for model learning, facilitating better learning of the embedded signals. Based on two state-of-the-art invisible offline-to-online messaging methods, we construct two datasets and augmentation strategies for training and testing localization models. Extensive experiments demonstrate the superior performance of the proposed DBDH over existing methods.
Title:
Semi-autonomous Robotic Disassembly Enhanced by Mixed Reality
Authors: Alireza Rastegarpanah, Cesar Alan Contreras, Rustam Stolkin
Abstract
In this study, we introduce "SARDiM," a modular semi-autonomous platform enhanced with mixed reality for industrial disassembly tasks. Through a case study focused on EV battery disassembly, SARDiM integrates Mixed Reality, object segmentation, teleoperation, force feedback, and variable autonomy. Utilising the ROS, Unity, and MATLAB platforms, alongside a joint impedance controller, SARDiM facilitates teleoperated disassembly. The approach combines FastSAM for real-time object segmentation, generating data which is subsequently processed through a cluster analysis algorithm to determine the centroid and orientation of the components, categorizing them by size and disassembly priority. This data guides the MoveIt platform in trajectory planning for the Franka Robot arm. SARDiM provides the capability to switch between two teleoperation modes: manual and semi-autonomous with variable autonomy. Each was evaluated using four different Interface Methods (IM): direct view, monitor feed, mixed reality with monitor feed, and point cloud mixed reality. Evaluations across the eight IMs demonstrated a 40.61% decrease in joint limit violations using Mode 2. Moreover, Mode 2-IM4 outperformed Mode 1-IM1 by achieving a 2.33%-time reduction while considerably increasing safety, making it optimal for operating in hazardous environments at a safe distance, with the same ease of use as teleoperation with a direct view of the environment.
Keyword: medical image segmentation
Title:
Implantable Adaptive Cells: differentiable architecture search to improve the performance of any trained U-shaped network
Authors: Emil Benedykciuk, Marcin Denkowski, Grzegorz Wójcik
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
This paper introduces a novel approach to enhance the performance of pre-trained neural networks in medical image segmentation using Neural Architecture Search (NAS) methods, specifically Differentiable Architecture Search (DARTS). We present the concept of Implantable Adaptive Cell (IAC), small but powerful modules identified through Partially-Connected DARTS, designed to be injected into the skip connections of an existing and already trained U-shaped model. Our strategy allows for the seamless integration of the IAC into the pre-existing architecture, thereby enhancing its performance without necessitating a complete retraining from scratch. The empirical studies, focusing on medical image segmentation tasks, demonstrate the efficacy of this method. The integration of specialized IAC cells into various configurations of the U-Net model increases segmentation accuracy by almost 2\% points on average for the validation dataset and over 3\% points for the training dataset. The findings of this study not only offer a cost-effective alternative to the complete overhaul of complex models for performance upgrades but also indicate the potential applicability of our method to other architectures and problem domains.
Keyword: unet
Title:
Spatio-Temporal SwinMAE: A Swin Transformer based Multiscale Representation Learner for Temporal Satellite Imagery
Abstract
Currently, the foundation models represented by large language models have made dramatic progress and are used in a very wide range of domains including 2D and 3D vision. As one of the important application domains of foundation models, earth observation has attracted attention and various approaches have been developed. When considering earth observation as a single image capture, earth observation imagery can be processed as an image with three or more channels, and when it comes with multiple image captures of different timestamps at one location, the temporal observation can be considered as a set of continuous image resembling video frames or medical SCAN slices. This paper presents Spatio-Temporal SwinMAE (ST-SwinMAE), an architecture which particularly focuses on representation learning for spatio-temporal image processing. Specifically, it uses a hierarchical Masked Auto-encoder (MAE) with Video Swin Transformer blocks. With the architecture, we present a pretrained model named Degas 100M as a geospatial foundation model. Also, we propose an approach for transfer learning with Degas 100M, which both pretrained encoder and decoder of MAE are utilized with skip connections added between them to achieve multi-scale information communication, forms an architecture named Spatio-Temporal SwinUNet (ST-SwinUNet). Our approach shows significant improvements of performance over existing state-of-the-art of foundation models. Specifically, for transfer learning of the land cover downstream task on the PhilEO Bench dataset, it shows 10.4\% higher accuracy compared with other geospatial foundation models on average.
Keyword: u-net
Title:
Continuous Learned Primal Dual
Authors: Christina Runkel, Ander Biguri, Carola-Bibiane Schönlieb
Subjects: Subjects:
Machine Learning (cs.LG); Image and Video Processing (eess.IV)
Abstract
Neural ordinary differential equations (Neural ODEs) propose the idea that a sequence of layers in a neural network is just a discretisation of an ODE, and thus can instead be directly modelled by a parameterised ODE. This idea has had resounding success in the deep learning literature, with direct or indirect influence in many state of the art ideas, such as diffusion models or time dependant models. Recently, a continuous version of the U-net architecture has been proposed, showing increased performance over its discrete counterpart in many imaging applications and wrapped with theoretical guarantees around its performance and robustness. In this work, we explore the use of Neural ODEs for learned inverse problems, in particular with the well-known Learned Primal Dual algorithm, and apply it to computed tomography (CT) reconstruction.
Title:
U-DiTs: Downsample Tokens in U-Shaped Diffusion Transformers
Authors: Yuchuan Tian, Zhijun Tu, Hanting Chen, Jie Hu, Chao Xu, Yunhe Wang
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Diffusion Transformers (DiTs) introduce the transformer architecture to diffusion tasks for latent-space image generation. With an isotropic architecture that chains a series of transformer blocks, DiTs demonstrate competitive performance and good scalability; but meanwhile, the abandonment of U-Net by DiTs and their following improvements is worth rethinking. To this end, we conduct a simple toy experiment by comparing a U-Net architectured DiT with an isotropic one. It turns out that the U-Net architecture only gain a slight advantage amid the U-Net inductive bias, indicating potential redundancies within the U-Net-style DiT. Inspired by the discovery that U-Net backbone features are low-frequency-dominated, we perform token downsampling on the query-key-value tuple for self-attention and bring further improvements despite a considerable amount of reduction in computation. Based on self-attention with downsampled tokens, we propose a series of U-shaped DiTs (U-DiTs) in the paper and conduct extensive experiments to demonstrate the extraordinary performance of U-DiT models. The proposed U-DiT could outperform DiT-XL/2 with only 1/6 of its computation cost. Codes are available at this https URL.
Title:
SalFAU-Net: Saliency Fusion Attention U-Net for Salient Object Detection
Authors: Kassaw Abraham Mulat, Zhengyong Feng, Tegegne Solomon Eshetie, Ahmed Endris Hasen
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
Salient object detection (SOD) remains an important task in computer vision, with applications ranging from image segmentation to autonomous driving. Fully convolutional network (FCN)-based methods have made remarkable progress in visual saliency detection over the last few decades. However, these methods have limitations in accurately detecting salient objects, particularly in challenging scenes with multiple objects, small objects, or objects with low resolutions. To address this issue, we proposed a Saliency Fusion Attention U-Net (SalFAU-Net) model, which incorporates a saliency fusion module into each decoder block of the attention U-net model to generate saliency probability maps from each decoder block. SalFAU-Net employs an attention mechanism to selectively focus on the most informative regions of an image and suppress non-salient regions. We train SalFAU-Net on the DUTS dataset using a binary cross-entropy loss function. We conducted experiments on six popular SOD evaluation datasets to evaluate the effectiveness of the proposed method. The experimental results demonstrate that our method, SalFAU-Net, achieves competitive performance compared to other methods in terms of mean absolute error (MAE), F-measure, s-measure, and e-measure.
Title:
Implantable Adaptive Cells: differentiable architecture search to improve the performance of any trained U-shaped network
Authors: Emil Benedykciuk, Marcin Denkowski, Grzegorz Wójcik
Subjects: Subjects:
Computer Vision and Pattern Recognition (cs.CV)
Abstract
This paper introduces a novel approach to enhance the performance of pre-trained neural networks in medical image segmentation using Neural Architecture Search (NAS) methods, specifically Differentiable Architecture Search (DARTS). We present the concept of Implantable Adaptive Cell (IAC), small but powerful modules identified through Partially-Connected DARTS, designed to be injected into the skip connections of an existing and already trained U-shaped model. Our strategy allows for the seamless integration of the IAC into the pre-existing architecture, thereby enhancing its performance without necessitating a complete retraining from scratch. The empirical studies, focusing on medical image segmentation tasks, demonstrate the efficacy of this method. The integration of specialized IAC cells into various configurations of the U-Net model increases segmentation accuracy by almost 2\% points on average for the validation dataset and over 3\% points for the training dataset. The findings of this study not only offer a cost-effective alternative to the complete overhaul of complex models for performance upgrades but also indicate the potential applicability of our method to other architectures and problem domains.
Title:
CCDM: Continuous Conditional Diffusion Models for Image Generation
Authors: Xin Ding, Yongwei Wang, Kao Zhang, Z. Jane Wang
Abstract
Continuous Conditional Generative Modeling (CCGM) aims to estimate the distribution of high-dimensional data, typically images, conditioned on scalar continuous variables known as regression labels. While Continuous conditional Generative Adversarial Networks (CcGANs) were initially designed for this task, their adversarial training mechanism remains vulnerable to extremely sparse or imbalanced data, resulting in suboptimal outcomes. To enhance the quality of generated images, a promising alternative is to replace CcGANs with Conditional Diffusion Models (CDMs), renowned for their stable training process and ability to produce more realistic images. However, existing CDMs encounter challenges when applied to CCGM tasks due to several limitations such as inadequate U-Net architectures and deficient model fitting mechanisms for handling regression labels. In this paper, we introduce Continuous Conditional Diffusion Models (CCDMs), the first CDM designed specifically for the CCGM task. CCDMs address the limitations of existing CDMs by introducing specially designed conditional diffusion processes, a modified denoising U-Net with a custom-made conditioning mechanism, a novel hard vicinal loss for model fitting, and an efficient conditional sampling procedure. With comprehensive experiments on four datasets with varying resolutions ranging from 64x64 to 192x192, we demonstrate the superiority of the proposed CCDM over state-of-the-art CCGM models, establishing new benchmarks in CCGM. Extensive ablation studies validate the model design and implementation configuration of the proposed CCDM. Our code is publicly available at this https URL.
Keyword: segmentation
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Title:
Keyword: medical image segmentation
Title:
Keyword: unet
Title:
Keyword: u-net
Title:
Title:
Title:
Title:
Title:
Keyword: interactive segmentation
There is no result