arXiv subjects

Junyu Chen

Publications and source records attributed to Junyu Chen.

At least 55 records · Page 3Linked to original sources

arXiv2025

Ambiguity-Aware and High-Order Relation Learning for Multi-Grained Image-Text Matching

Image-text matching is crucial for bridging the semantic gap between computer vision and natural language processing. However, existing methods still face challenges in handling high-order associations and semantic ambiguities among similar instances. These ambiguities arise from subtle differences between soft positive samples (semantically similar but incorrectly labeled) and soft negative samples (locally matched but globally inconsistent), creating matching uncertainties. Furthermore, current methods fail to fully utilize the neighborhood relationships among semantically similar instances within training batches, limiting the model's ability to learn high-order shared knowledge. This paper proposes the Ambiguity-Aware and High-order Relation learning framework (AAHR) to address these issues. AAHR constructs a unified representation space through dynamic clustering prototype contrastive learning, effectively mitigating the soft positive sample problem. The framework introduces global and local feature extraction mechanisms and an adaptive aggregation network, significantly enhancing full-grained semantic understanding capabilities. Additionally, AAHR employs intra-modal and inter-modal correlation matrices to investigate neighborhood relationships among sample instances thoroughly. It incorporates GNN to enhance semantic interactions between instances. Furthermore, AAHR integrates momentum contrastive learning to expand the negative sample set. These combined strategies significantly improve the model's ability to discriminate between features. Experimental results demonstrate that AAHR outperforms existing state-of-the-art methods on Flickr30K, MSCOCO, and ECCV Caption datasets, considerably improving the accuracy and efficiency of image-text matching. The code and model checkpoints for this research are available at https://github.com/Image-Text-Matching/AAHR .

cs.CV↗

arXiv2025

Visual Semantic Description Generation with MLLMs for Image-Text Matching

Image-text matching (ITM) aims to address the fundamental challenge of aligning visual and textual modalities, which inherently differ in their representations, continuous, high-dimensional image features vs. discrete, structured text. We propose a novel framework that bridges the modality gap by leveraging multimodal large language models (MLLMs) as visual semantic parsers. By generating rich Visual Semantic Descriptions (VSD), MLLMs provide semantic anchor that facilitate cross-modal alignment. Our approach combines: (1) Instance-level alignment by fusing visual features with VSD to enhance the linguistic expressiveness of image representations, and (2) Prototype-level alignment through VSD clustering to ensure category-level consistency. These modules can be seamlessly integrated into existing ITM models. Extensive experiments on Flickr30K and MSCOCO demonstrate substantial performance improvements. The approach also exhibits remarkable zero-shot generalization to cross-domain tasks, including news and remote sensing ITM. The code and model checkpoints are available at https://github.com/Image-Text-Matching/VSD.

cs.MM↗

arXiv2025

DC-AR: Efficient Masked Autoregressive Image Generation with Deep Compression Hybrid Tokenizer

We introduce DC-AR, a novel masked autoregressive (AR) text-to-image generation framework that delivers superior image generation quality with exceptional computational efficiency. Due to the tokenizers' limitations, prior masked AR models have lagged behind diffusion models in terms of quality or efficiency. We overcome this limitation by introducing DC-HT - a deep compression hybrid tokenizer for AR models that achieves a 32x spatial compression ratio while maintaining high reconstruction fidelity and cross-resolution generalization ability. Building upon DC-HT, we extend MaskGIT and create a new hybrid masked autoregressive image generation framework that first produces the structural elements through discrete tokens and then applies refinements via residual tokens. DC-AR achieves state-of-the-art results with a gFID of 5.49 on MJHQ-30K and an overall score of 0.69 on GenEval, while offering 1.5-7.9x higher throughput and 2.0-3.5x lower latency compared to prior leading diffusion and autoregressive models.

cs.CV↗

arXiv2025

Pretraining Deformable Image Registration Networks with Random Images

Recent advances in deep learning-based medical image registration have shown that training deep neural networks~(DNNs) does not necessarily require medical images. Previous work showed that DNNs trained on randomly generated images with carefully designed noise and contrast properties can still generalize well to unseen medical data. Building on this insight, we propose using registration between random images as a proxy task for pretraining a foundation model for image registration. Empirical results show that our pretraining strategy improves registration accuracy, reduces the amount of domain-specific data needed to achieve competitive performance, and accelerates convergence during downstream training, thereby enhancing computational efficiency.

cs.CV↗

arXiv2025

Deep Compression Autoencoder for Efficient High-Resolution Diffusion Models

We present Deep Compression Autoencoder (DC-AE), a new family of autoencoder models for accelerating high-resolution diffusion models. Existing autoencoder models have demonstrated impressive results at a moderate spatial compression ratio (e.g., 8x), but fail to maintain satisfactory reconstruction accuracy for high spatial compression ratios (e.g., 64x). We address this challenge by introducing two key techniques: (1) Residual Autoencoding, where we design our models to learn residuals based on the space-to-channel transformed features to alleviate the optimization difficulty of high spatial-compression autoencoders; (2) Decoupled High-Resolution Adaptation, an efficient decoupled three-phases training strategy for mitigating the generalization penalty of high spatial-compression autoencoders. With these designs, we improve the autoencoder's spatial compression ratio up to 128 while maintaining the reconstruction quality. Applying our DC-AE to latent diffusion models, we achieve significant speedup without accuracy drop. For example, on ImageNet 512x512, our DC-AE provides 19.1x inference speedup and 17.9x training speedup on H100 GPU for UViT-H while achieving a better FID, compared with the widely used SD-VAE-f8 autoencoder. Our code is available at https://github.com/mit-han-lab/efficientvit.

cs.CV↗

arXiv2025

SANA 1.5: Efficient Scaling of Training-Time and Inference-Time Compute in Linear Diffusion Transformer

This paper presents SANA-1.5, a linear Diffusion Transformer for efficient scaling in text-to-image generation. Building upon SANA-1.0, we introduce three key innovations: (1) Efficient Training Scaling: A depth-growth paradigm that enables scaling from 1.6B to 4.8B parameters with significantly reduced computational resources, combined with a memory-efficient 8-bit optimizer. (2) Model Depth Pruning: A block importance analysis technique for efficient model compression to arbitrary sizes with minimal quality loss. (3) Inference-time Scaling: A repeated sampling strategy that trades computation for model capacity, enabling smaller models to match larger model quality at inference time. Through these strategies, SANA-1.5 achieves a text-image alignment score of 0.81 on GenEval, which can be further improved to 0.96 through inference scaling with VILA-Judge, establishing a new SoTA on GenEval benchmark. These innovations enable efficient model scaling across different compute budgets while maintaining high quality, making high-quality image generation more accessible. Our code and pre-trained models are released.

cs.CV↗

arXiv2025

Correlation Ratio for Unsupervised Learning of Multi-modal Deformable Registration

In recent years, unsupervised learning for deformable image registration has been a major research focus. This approach involves training a registration network using pairs of moving and fixed images, along with a loss function that combines an image similarity measure and deformation regularization. For multi-modal image registration tasks, the correlation ratio has been a widely-used image similarity measure historically, yet it has been underexplored in current deep learning methods. Here, we propose a differentiable correlation ratio to use as a loss function for learning-based multi-modal deformable image registration. This approach extends the traditionally non-differentiable implementation of the correlation ratio by using the Parzen windowing approximation, enabling backpropagation with deep neural networks. We validated the proposed correlation ratio on a multi-modal neuroimaging dataset. In addition, we established a Bayesian training framework to study how the trade-off between the deformation regularizer and similarity measures, including mutual information and our proposed correlation ratio, affects the registration performance.

eess.IV↗

arXiv2025

VILA-U: a Unified Foundation Model Integrating Visual Understanding and Generation

VILA-U is a Unified foundation model that integrates Video, Image, Language understanding and generation. Traditional visual language models (VLMs) use separate modules for understanding and generating visual content, which can lead to misalignment and increased complexity. In contrast, VILA-U employs a single autoregressive next-token prediction framework for both tasks, eliminating the need for additional components like diffusion models. This approach not only simplifies the model but also achieves near state-of-the-art performance in visual language understanding and generation. The success of VILA-U is attributed to two main factors: the unified vision tower that aligns discrete visual tokens with textual inputs during pretraining, which enhances visual perception, and autoregressive image generation can achieve similar quality as diffusion models with high-quality dataset. This allows VILA-U to perform comparably to more complex models using a fully token-based autoregressive framework.

cs.CV↗

arXiv2025

Cross-Attention Fusion of MRI and Jacobian Maps for Alzheimer's Disease Diagnosis

Early diagnosis of Alzheimer's disease (AD) is critical for intervention before irreversible neurodegeneration occurs. Structural MRI (sMRI) is widely used for AD diagnosis, but conventional deep learning approaches primarily rely on intensity-based features, which require large datasets to capture subtle structural changes. Jacobian determinant maps (JSM) provide complementary information by encoding localized brain deformations, yet existing multimodal fusion strategies fail to fully integrate these features with sMRI. We propose a cross-attention fusion framework to model the intrinsic relationship between sMRI intensity and JSM-derived deformations for AD classification. Using the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset, we compare cross-attention, pairwise self-attention, and bottleneck attention with four pre-trained 3D image encoders. Cross-attention fusion achieves superior performance, with mean ROC-AUC scores of 0.903 (+/-0.033) for AD vs. cognitively normal (CN) and 0.692 (+/-0.061) for mild cognitive impairment (MCI) vs. CN. Despite its strong performance, our model remains highly efficient, with only 1.56 million parameters--over 40 times fewer than ResNet-34 (63M) and Swin UNETR (61.98M). These findings demonstrate the potential of cross-attention fusion for improving AD diagnosis while maintaining computational efficiency.

eess.IV↗

arXiv2025

MobileH2R: Learning Generalizable Human to Mobile Robot Handover Exclusively from Scalable and Diverse Synthetic Data

This paper introduces MobileH2R, a framework for learning generalizable vision-based human-to-mobile-robot (H2MR) handover skills. Unlike traditional fixed-base handovers, this task requires a mobile robot to reliably receive objects in a large workspace enabled by its mobility. Our key insight is that generalizable handover skills can be developed in simulators using high-quality synthetic data, without the need for real-world demonstrations. To achieve this, we propose a scalable pipeline for generating diverse synthetic full-body human motion data, an automated method for creating safe and imitation-friendly demonstrations, and an efficient 4D imitation learning method for distilling large-scale demonstrations into closed-loop policies with base-arm coordination. Experimental evaluations in both simulators and the real world show significant improvements (at least +15% success rate) over baseline methods in all cases. Experiments also validate that large-scale and diverse synthetic data greatly enhances robot learning, highlighting our scalable framework.

cs.RO↗

arXiv2024

A survey on deep learning in medical image registration: new technologies, uncertainty, evaluation metrics, and beyond

Deep learning technologies have dramatically reshaped the field of medical image registration over the past decade. The initial developments, such as regression-based and U-Net-based networks, established the foundation for deep learning in image registration. Subsequent progress has been made in various aspects of deep learning-based registration, including similarity measures, deformation regularizations, network architectures, and uncertainty estimation. These advancements have not only enriched the field of image registration but have also facilitated its application in a wide range of tasks, including atlas construction, multi-atlas segmentation, motion estimation, and 2D-3D registration. In this paper, we present a comprehensive overview of the most recent advancements in deep learning-based image registration. We begin with a concise introduction to the core concepts of deep learning-based image registration. Then, we delve into innovative network architectures, loss functions specific to registration, and methods for estimating registration uncertainty. Additionally, this paper explores appropriate evaluation metrics for assessing the performance of deep learning models in registration tasks. Finally, we highlight the practical applications of these novel techniques in medical imaging and discuss the future prospects of deep learning-based image registration.

eess.IV↗

arXiv2024

Gene-Metabolite Association Prediction with Interactive Knowledge Transfer Enhanced Graph for Metabolite Production

In the rapidly evolving field of metabolic engineering, the quest for efficient and precise gene target identification for metabolite production enhancement presents significant challenges. Traditional approaches, whether knowledge-based or model-based, are notably time-consuming and labor-intensive, due to the vast scale of research literature and the approximation nature of genome-scale metabolic model (GEM) simulations. Therefore, we propose a new task, Gene-Metabolite Association Prediction based on metabolic graphs, to automate the process of candidate gene discovery for a given pair of metabolite and candidate-associated genes, as well as presenting the first benchmark containing 2474 metabolites and 1947 genes of two commonly used microorganisms Saccharomyces cerevisiae (SC) and Issatchenkia orientalis (IO). This task is challenging due to the incompleteness of the metabolic graphs and the heterogeneity among distinct metabolisms. To overcome these limitations, we propose an Interactive Knowledge Transfer mechanism based on Metabolism Graph (IKT4Meta), which improves the association prediction accuracy by integrating the knowledge from different metabolism graphs. First, to build a bridge between two graphs for knowledge transfer, we utilize Pretrained Language Models (PLMs) with external knowledge of genes and metabolites to help generate inter-graph links, significantly alleviating the impact of heterogeneity. Second, we propagate intra-graph links from different metabolic graphs using inter-graph links as anchors. Finally, we conduct the gene-metabolite association prediction based on the enriched metabolism graphs, which integrate the knowledge from multiple microorganisms. Experiments on both types of organisms demonstrate that our proposed methodology outperforms baselines by up to 12.3% across various link prediction frameworks.

cs.AI↗

arXiv2024

SANA: Efficient High-Resolution Image Synthesis with Linear Diffusion Transformers

We introduce Sana, a text-to-image framework that can efficiently generate images up to 4096$\times$4096 resolution. Sana can synthesize high-resolution, high-quality images with strong text-image alignment at a remarkably fast speed, deployable on laptop GPU. Core designs include: (1) Deep compression autoencoder: unlike traditional AEs, which compress images only 8$\times$, we trained an AE that can compress images 32$\times$, effectively reducing the number of latent tokens. (2) Linear DiT: we replace all vanilla attention in DiT with linear attention, which is more efficient at high resolutions without sacrificing quality. (3) Decoder-only text encoder: we replaced T5 with modern decoder-only small LLM as the text encoder and designed complex human instruction with in-context learning to enhance the image-text alignment. (4) Efficient training and sampling: we propose Flow-DPM-Solver to reduce sampling steps, with efficient caption labeling and selection to accelerate convergence. As a result, Sana-0.6B is very competitive with modern giant diffusion model (e.g. Flux-12B), being 20 times smaller and 100+ times faster in measured throughput. Moreover, Sana-0.6B can be deployed on a 16GB laptop GPU, taking less than 1 second to generate a 1024$\times$1024 resolution image. Sana enables content creation at low cost. Code and model will be publicly released.

cs.CV↗

arXiv2024

Semidefinite Relaxations of the Gromov-Wasserstein Distance

The Gromov-Wasserstein (GW) distance is an extension of the optimal transport problem that allows one to match objects between incomparable spaces. At its core, the GW distance is specified as the solution of a non-convex quadratic program and is not known to be tractable to solve. In particular, existing solvers for the GW distance are only able to find locally optimal solutions. In this work, we propose a semi-definite programming (SDP) relaxation of the GW distance. The relaxation can be viewed as the Lagrangian dual of the GW distance augmented with constraints that relate to the linear and quadratic terms of transportation plans. In particular, our relaxation provides a tractable (polynomial-time) algorithm to compute globally optimal transportation plans (in some instances) together with an accompanying proof of global optimality. Our numerical experiments suggest that the proposed relaxation is strong in that it frequently computes the globally optimal solution. Our Python implementation is available at https://github.com/tbng/gwsdp.

math.OC↗

arXiv2024

Exactness Conditions for Semidefinite Relaxations of the Quadratic Assignment Problem

The Quadratic Assignment Problem (QAP) is an important discrete optimization instance that encompasses many well-known combinatorial optimization problems, and has applications in a wide range of areas such as logistics and computer vision. The QAP, unfortunately, is NP-hard to solve. To address this difficulty, a number of semidefinite relaxations of the QAP have been developed. These techniques are known to be powerful in that they compute globally optimal solutions in many instances, and are often deployed as sub-routines within enumerative procedures for solving QAPs. In this paper, we investigate the strength of these semidefinite relaxations. Our main result is a deterministic set of conditions on the input matrices -- specified via linear inequalities -- under which these semidefinite relaxations are exact. Our result is simple to state, in the hope that it serves as a foundation for subsequent studies on semidefinite relaxations of the QAP as well as other related combinatorial optimization problems. As a corollary, we show that the semidefinite relaxation is exact under a perturbation model whereby the input matrices differ by a suitably small perturbation term. One technical difficulty we encounter is that the set of dual feasible solutions is not closed. To circumvent these difficulties, our analysis constructs a sequence of dual feasible solutions whose objective value approaches the primal objective, but never attains it.

math.OC↗

arXiv2024

HART: Efficient Visual Generation with Hybrid Autoregressive Transformer

We introduce Hybrid Autoregressive Transformer (HART), an autoregressive (AR) visual generation model capable of directly generating 1024x1024 images, rivaling diffusion models in image generation quality. Existing AR models face limitations due to the poor image reconstruction quality of their discrete tokenizers and the prohibitive training costs associated with generating 1024px images. To address these challenges, we present the hybrid tokenizer, which decomposes the continuous latents from the autoencoder into two components: discrete tokens representing the big picture and continuous tokens representing the residual components that cannot be represented by the discrete tokens. The discrete component is modeled by a scalable-resolution discrete AR model, while the continuous component is learned with a lightweight residual diffusion module with only 37M parameters. Compared with the discrete-only VAR tokenizer, our hybrid approach improves reconstruction FID from 2.11 to 0.30 on MJHQ-30K, leading to a 31% generation FID improvement from 7.85 to 5.38. HART also outperforms state-of-the-art diffusion models in both FID and CLIP score, with 4.5-7.7x higher throughput and 6.9-13.4x lower MACs. Our code is open sourced at https://github.com/mit-han-lab/hart.

cs.CV↗

arXiv2024

Unsupervised Learning of Multi-modal Affine Registration for PET/CT

Affine registration plays a crucial role in PET/CT imaging, where aligning PET with CT images is challenging due to their respective functional and anatomical representations. Despite the significant promise shown by recent deep learning (DL)-based methods in various medical imaging applications, their application to multi-modal PET/CT affine registration remains relatively unexplored. This study investigates a DL-based approach for PET/CT affine registration. We introduce a novel method using Parzen windowing to approximate the correlation ratio, which acts as the image similarity measure for training DNNs in multi-modal registration. Additionally, we propose a multi-scale, instance-specific optimization scheme that iteratively refines the DNN-generated affine parameters across multiple image resolutions. Our method was evaluated against the widely used mutual information metric and a popular optimization-based technique from the ANTs package, using a large public FDG-PET/CT dataset with synthetic affine transformations. Our approach achieved a mean Dice Similarity Coefficient (DSC) of 0.870, outperforming the compared methods and demonstrating its effectiveness in multi-modal PET/CT image registration.

eess.IV↗

arXiv2024

RSEA-MVGNN: Multi-View Graph Neural Network with Reliable Structural Enhancement and Aggregation

Graph Neural Networks (GNNs) have exhibited remarkable efficacy in learning from multi-view graph data. In the framework of multi-view graph neural networks, a critical challenge lies in effectively combining diverse views, where each view has distinct graph structure features (GSFs). Existing approaches to this challenge primarily focus on two aspects: 1) prioritizing the most important GSFs, 2) utilizing GNNs for feature aggregation. However, prioritizing the most important GSFs can lead to limited feature diversity, and existing GNN-based aggregation strategies equally treat each view without considering view quality. To address these issues, we propose a novel Multi-View Graph Neural Network with Reliable Structural Enhancement and Aggregation (RSEA-MVGNN). Firstly, we estimate view-specific uncertainty employing subjective logic. Based on this uncertainty, we design reliable structural enhancement by feature de-correlation algorithm. This approach enables each enhancement to focus on different GSFs, thereby achieving diverse feature representation in the enhanced structure. Secondly, the model learns view-specific beliefs and uncertainty as opinions, which are utilized to evaluate view quality. Based on these opinions, the model enables high-quality views to dominate GNN aggregation, thereby facilitating representation learning. Experimental results conducted on five real-world datasets demonstrate that RSEA-MVGNN outperforms several state-of-the-art GNN-based methods.

cs.LG↗