Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks129 from public data
- 531
- RMT: Retentive Networks Meet Vision Transformers270
This work extends the RetNet's temporal decay mechanism to the spatial do-main, and proposes a spatial decay matrix based on the Manhattan distance to introduce the explicit spatial prior to Self-Attention, aiming to reduce the computational burden of modeling global information without disrupting the spa-tial decay matrix.
- 151
- 148
- 124
- Vision Transformer with Super Token Sampling122
A simple yet strong super token attention (STA) mechanism with three steps: the first samples super tokens from visual tokens via sparse association learning, the second performs self-attention on super tokens, and the last maps them back to the original token space.
- Disentangled Variational Representation for Heterogeneous Face Recognition97
This paper model a face representation with an intrinsic identity information and its within-person variations and proposes an alternative optimization scheme for the disentangled variational representation part and the heterogeneous face recognition network part that effectively reduces the NIR and VIS domain discrepancy and alleviates over-fitting.
- Wavelet Domain Generative Adversarial Network for Multi-scale Face Hallucination91
This paper proposes a wavelet-domain generative adversarial method that can ultra-resolve a very low-resolution face image and can significantly improve identification accuracy for low-resolution face images captured in the wild.
- MMFakeBench: A Mixed-Source Multimodal Misinformation Detection Benchmark for LVLMs83
MMFakeBench is introduced, the first comprehensive benchmark for mixed-source MMD, and MMD-Agent is proposed, a novel approach to integrate the reasoning, action, and tool-use capabilities of LVLM agents, significantly enhancing accuracy and generalization.
- FKA-Owl: Advancing Multimodal Fake News Detection through Knowledge-Augmented LVLMs81
This paper proposes FKA-Owl, a novel framework that leverages forgery-specific knowledge to augment LVLMs, enabling them to reason about manipulations effectively and achieves superior cross-domain performance compared to previous methods.
- A Survey of Deep Facial Attribute Analysis72
A general pipeline that deep facial attribute analysis follows, which comprises two stages: data preprocessing and model construction is summarized and the underlying theories of this two-stage pipeline for both FAE and FAM are introduced.
- 71
- Rethinking Local Perception in Lightweight Vision Transformer66
CloFormer is introduced, a lightweight vision transformer that leverages context-aware local enhancement and introduces AttnConv, a convolution operator in attention's style that enables the model to perceive high-frequency and low-frequency information.
- Arbitrary Talking Face Generation via Attentional Audio-Visual Coherence Learning65
A novel arbitrary talking face generation framework is proposed by discovering the audio-visual coherence via the proposed Asymmetric Mutual Information Estimator (AMIE) and a Dynamic Attention (DA) block by selectively focusing the lip area of the input image during the training stage, to further enhance lip synchronization.
- 64
- Selective Wavelet Attention Learning for Single Image Deraining60
A selectiveWavelet attention learning method by learning a series of wavelet attention maps to guide the separation of rain and background information in both spatial and frequency domains that achieves more appealing results both quantitatively and qualitatively than the recent state-of-the-art methods.
- Breaking the Low-Rank Dilemma of Linear Attention54
Rank-Augmented Linear Attention (RALA), which rivals the performance of Softmax attention while maintaining linear complexity and high efficiency, is introduced and the Rank-Augmented Vision Linear Transformer (RAVLT) is constructed.
- 49
- 47
- 43
- Jailbreak Attacks and Defenses against Multimodal Generative Models: A Survey42
A detailed taxonomy of attack methods, defense mechanisms, and evaluation frameworks specific to multimodal generative models is presented, including modalities such as Any-to-Text, Any-to-Vision, and Any-to-Any within generative systems.
- LAMP-HQ: A Large-Scale Multi-pose High-Quality Database and Benchmark for NIR-VIS Face Recognition42
A benchmark is furnished along with the protocol for NIR-VIS face recognition via generation on LAMP-HQ and a spectral approach is proposed to reduce the domain performance to improve the performance of the model training process.
- Learning Disentangled Representation for Cross-Modal Retrieval with Deep Mutual Information Estimation40
Extensive experiments on two large-scale benchmark datasets demonstrate that the proposed method can learn better visual-semantic embedding and achieve state-of-the-art cross-modal retrieval results.
- Heterogeneous Test-Time Training for Multi-Modal Person Re-identification36
A Cross-identity Inter-modal Margin (CIM) loss is proposed to amplify the differentiation among distinct identity samples and a Multi-modal Test-time Training (MTT) strategy is designed to enhance the generalization of the model by leveraging the relationships in the heterogeneous modalities and the information existing in the test data.
- ViTAR: Vision Transformer with Any Resolution36
This work proposes a novel module for dynamic resolution adjustment, designed with a single Transformer block, specifically to achieve highly efficient incremental token integration and introduces fuzzy positional encoding in the Vision Transformer to provide consistent positional awareness across multiple resolutions, thereby preventing overfitting to any single training resolution.
- 35
- 30
- Pluralistic Aging Diffusion Autoencoder29
A novel CLIP-driven Pluralistic Aging Diffusion Autoencoder (PADA) to enhance the diversity of aging patterns and presents Probabilistic Aging Embedding (PAE) to capture diverse high-level aging patterns, which represents age information as probabilistic distributions in the common CLIP latent space.
- 29
- 28
- Image-matching as a medical diagnostic support tool (DST) for brain diseases in children28
A novel diagnostic support tool based on medical image matching using MR brain images was developed and evaluated and it was found that in most cases, the image-matching method was able to quickly retrieve images with relevant diagnostic content.
- LoRA-IR: Taming Low-Rank Experts for Efficient All-in-One Image Restoration27
LoRA-IR is a flexible framework that dynamically leverages compact low-rank experts to facilitate efficient all-in-one image restoration and significantly enhances the model's adaptability to diverse and unknown degradations in complex real-world scenarios.
- 26
- 25
- 23
- 23
- Parallel Augmentation and Dual Enhancement for Occluded Person Re-Identification22
This work designs a parallel augmentation mechanism (PAM) to generate more suitable occluded data to mitigate the negative effects of unbalanced data and proposes the global and local dual enhancement strategy (DES) to promote the context information and details.
- Disentangled Representation Learning of Makeup Portraits in the Wild22
Both visual and quantitative experimental results on four makeup datasets demonstrate the superiority of the proposed disentangled feature learning strategy, which proposes a cosmetic loss to learn makeup styles in a delicate way.
- 21
- High-Resolution Talking Face Generation via Mutual Information Approximation.21
A novel high-resolution talking face generation model for arbitrary person is proposed by discovering the cross-modality coherence via Mutual Information Approximation (MIA) by assuming the modality difference between audio and video is larger that of real video and generated video.
- 20
- Contrastive attention network with dense field estimation for face completion20
A self-supervised Siamese inference network is proposed to improve the generalization and robustness of encoders and improves the performance of masked face recognition dramatically.
- 19
- Uncertainty-Aware Source-Free Adaptive Image Super-Resolution with Wavelet Augmentation Transformer18
A novel wavelet-based augmentation method, named Wavelet Augmentation Transformer (WAT), which can be flexibly incorporated with existing networks, to implicitly produce useful augmented data, and several regularization losses are proposed to constrain target LR and SR images in the frequency domain.
- Portrait Diffusion: Training-free Face Stylization with Chain-of-Painting16
This paper proposes a training-free face stylization framework, named Portrait Diffusion, which leverages off-the-shelf text-to-image diffusion models, eliminating the need for fine-tuning specific examples.
- Video-Teller: Enhancing Cross-Modal Generation with Fusion and Decoupling16
Experimental results demonstrate the efficacy of the proposed video-language foundation model in accurately comprehending videos and generating coherent and precise language descriptions and the fine-grained alignment enhances the model's capabilities with only 13% extra parameters in training and zero additional cost in inference.
- Vision Transformer with Sparse Scan Prior15
This work proposes a Sparse Scan Self-Attention mechanism (S3A) that mirrors the human eye's functionality and significantly reduces the computational load of vision models, and introduces the Sparse Scan Vision Transformer (SSViT).
- Test-time Forgery Detection with Spatial-Frequency Prompt Learning14
A Dynamic Dual-spectrum Interaction Network that allows test-time training with uncertainty guidance and spatial-frequency prompt learning and proposes a spatial-frequency prompt learning method to effectively enhance the generalization of the forgery detection model.
- BLAN: Bi-directional ladder attentive network for facial attribute prediction13
A novel Bi-directional Ladder Attentive Network (BLAN) to learn hierarchical representations, covering the correlations between feature hierarchies and attribute characteristics, and outperforms state-of-the-art methods.
- Image database for digital hand atlas13
A new digital hand atlas with a large set of clinically normal hand images of diverse ethnic groups built on a large image database and current Internet technology provides an alternative to supplement or replace the traditional one for a quantitative, accurate and cost-effective assessment of bone age.
- ZePo: Zero-Shot Portrait Stylization with Faster Sampling10
This paper presents an inversion-free portrait stylization framework based on diffusion models that accomplishes content and style feature fusion in merely four sampling steps, and introduces a Style Enhancement Attention Control technique that meticulously merges content and style features within the attention space of the target image.
- Informative Sample Mining Network for Multi-domain Image-to-Image Translation10
It is revealed that improving the sample selection strategy is an effective solution for multi-domain image-to-image translation, and a novel multi-stage sample training scheme to reduce sample hardness while preserving sample informativeness is proposed.
- Unlocking the Potential of Difficulty Prior in RL-based Multimodal Reasoning9
This work investigates how explicitly modeling problem's difficulty prior information shapes the effectiveness of reinforcement learning based fine-tuning for multimodal reasoning, and implements an online advantage differentiation to adaptively reweight advantages estimation.
- 9
- Free-Form Image Inpainting via Contrastive Attention Network9
A self-supervised Siamese inference network that can encode contextual semantics from full resolution images and obtain more discriminative representations and a multi-scale decoder with a novel dual attention fusion module (DAF), which can combine both the restored and known regions in a smooth way.
- Content-based image retrieval in picture archiving and communication systems9
A CBIR system that is integrated with PACS and RIS, by using a user-supplied query image to retrieve similar images fromPACS and get corresponding reports from RIS, and employs ACR index for radiological diagnosis to reduce the search space and to provide meaningful results.
- 8
- DeVAn: Dense Video Annotation for Video-Language Models8
A novel human annotated dataset for evaluating the ability for visual-language models to generate both short and long descriptions for real-world video clips, termed DeVAn (Dense Video Annotation), and it is found that model-based evaluation metrics provide more semantically-oriented and human-aligned evaluation.
- 8
- 8
- Semantic Equitable Clustering: A Simple and Effective Strategy for Clustering Vision Tokens7
This work introduces a fast and balanced clustering method, named Semantic Equitable Clustering (SEC), which clusters tokens based on their global semantic relevance in an efficient, straightforward manner and proposes a versatile vision backbone, SECViT.
- 7
- 7
- 7
- 6
- LAMP-HQ: A Large-Scale Multi-Pose High-Quality Database for NIR-VIS Face Recognition.6
A novel exemplar-based variational spectral attention network to produce high-fidelity VIS images from NIR data and a spectral conditional attention module is introduced to reduce the domain gap between NIR and VIS data and improve the performance of NIR-VIS heterogeneous face recognition on various databases including the LAMP-HQ.
- Video Forgery Detection Using Spatio-Temporal Dual Transformer5
Extensive experiments show that the proposed framework for video forgery detection has high accuracy and generalization, outperforming the current state-of-the-art methods.
- 5
- 4
- 4
- 4
- LightCvT: Audio Forgery Detection via Fusion of Light CNN and Transformer4
A light transformer-based network LightCvT is proposed, which attempts to capture both local feature and non-local long-range dependencies to improve the detection accuracy of audio forgery and introduces the self-attention mechanism into the convolutional token embedding to further model the relations between local regions.
- 4
- Experiences With a Prototype Tracking and Verification System Implemented Within an Imaging Center4
The LTVS enabled health care providers to streamline the patient workflow, protect against erroneous examinations and create a security zone to prevent, and audit unauthorized access to patient health care data required by the Health Insurance Portability and Accountability Act mandate.
- Variational Capsules for Image Analysis and Synthesis3
The capsule is extended to a generative version, named variational capsules (VCs), where each VC produces a latent variable for a specific entity, making it possible to integrate image analysis and image synthesis into a unified framework.
- 3
- Uncertainty-Guided Test-Time Training for Face Forgery Detection3
A Dynamic Dual-spectrum Interaction Network (DDIN) that allows test-time training with uncertainty guidance and can dynamically fuse the features with quality discrepancies, thus improving the generalization of forgery detection.
- 3
- 3
- 3
- 3
- Data mining and visualization of average images in a digital hand atlas3
The "average" reference image in the digital atlas is selected for each of the groups of normal developed children with the best representative skeletal maturity based on bony features, to provide clinicians accurate reference images to aid the bone age assessment process.
- 2
- T^2Agent A Tool-augmented Multimodal Misinformation Detection Agent with Monte Carlo Tree Search2
T2Agent is a novel misinformation detection agent that incorporates an extensible toolkit with Monte Carlo Tree Search (MCTS), and extends traditional MCTS with multi-source verification, which decomposes the task into coordinated subtasks targeting different forgery sources.
- ID-Cloak: Crafting Identity-Specific Cloaks Against Personalized Text-to-Image Generation2
This work is the first to investigate the creation of identity-specific cloaks (ID-Cloak) that safeguard all images belong to a specific identity, and believes this method marks a notable advance in realistic privacy protection.
- DiffMAC: Diffusion Manifold Hallucination Correction for High Generalization Blind Face Restoration2
A Diffusion-Information-Diffusion (DID) framework to tackle diffusion manifold hallucination correction (DiffMAC), which achieves high-generalization face restoration in diverse degraded scenes and heterogeneous domains and demonstrates the superiority of DiffMAC over state-of-the-art methods.
- 2
- 2
- 2
- 2
- Style-based Variational Autoencoder for Real-World Super-Resolution.2
This paper proposes a novel Style-based Super-Resolution Variational Autoencoder network (SSRVAE) that contains a style Variationalautoencoding (styleVAE), a SR Network, and a mutual information estimation (MI) to get better style information.
- HIPAA-compliant automatic monitoring system for RIS-integrated PACS operation2
The design of a HIPAA-compliant automatic monitoring system for RIS-Integrated PACS Operation is presented, and the preliminary results performed by this monitoring system on a clinical RIS-integratedPACS is given.
- Technical experiences of implementing a wireless tracking and facial biometric verification system for a clinical environment2
The technical design and implementation experiences encountered during the development of a Location Tracking and Verification System (LTVS) for a clinical environment are presented.
- Image matching as a diagnostic support tool2
It is found that the image-matching method is able to retrieve images with relevant diagnostic content, making it highly attractive as a diagnostic support tool for pediatric brain diseases.
- 1
- 1
- NOFT: Test-Time Noise Finetune via Information Bottleneck for Highly Correlated Asset Creation1
A plug-and-play noise finetune NOFT module employed by Stable Diffusion to generate highly correlated and diverse images and demonstrates that NOFT is a powerful general reimagine approach to efficiently fine-tune the 2D/3D AIGC assets with text or image guidance.
- 1
- 1
- 1
- 1
- Heterogeneous Facial Analysis and Synthesis1
This chapter gives an overview of the heterogeneous problem in facial analysis as well as its synthesis solution, followed by a brief outline of the rest chapters.
- 1
- 1
- Development of an embedded instrument for autofocus and polarization alignment of polarization maintaining fiber1
The development and implementation of a practical instrument based on an embedded technique for autofocus and polarization alignment of polarization maintaining fiber and various image-based algorithms of alignment detection were developed with high calculation speed and strong robustness.
- Knowledge-based approach for functional MRI analysis by SOM neural network using prior labels from Talairach stereotaxic space1
The purpose of this approach is to avoid the inclusion of baseline voxels in the training of the SOM neural network by using three different levels of segmentation and mapping from Talairach space: (1) voxel partitions at the lobe level, (2)voxel partition at the gyrus level and (3) v oxel partitionsat the cell level (Brodmann areas).
- Advancing Vision Transformer With Enhanced Spatial Priors–
Euclidean enhanced Vision Transformer (EVT) is an expanded version that incorporates several key improvements and offers a more sophisticated and adaptable approach to incorporating spatial priors into the Self-Attention mechanism, thus overcoming some of the limitations associated with RMT and further enhancing its applicability in various computer vision tasks.
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- Comparison of fingerprint and facial biometric verification technologies for user access and patient identification in a clinical environment–
This work installs two biometric identification systems including fingerprint and facial recognition systems at an outpatient imaging facility, Healthcare Consultation Center II (HCCII), and evaluated each solution and documented the advantages and pitfalls of each biometric technology in this clinical environment.
- –
- –
- Building a digital atlas from a normal hand-image database–
This paper presents an atlasing method that will automatically select the reference images from the normal collection database to form a digital atlas that can always be replenished with new images reflecting more accurately the current population and skeletal development.
- –
- Develop security architecture for both in-house healthcare information systems and electronic patient record–
Two security components, certificate authoring (CA) system and patient record digital signature management (DSPR) system, as well as electronic envelope technology, are introduced into the current hospital healthcare information infrastructure to provide security measures and functions.
- Computed radiographic image post-processing for automatic optimal display in picture archiving and communication system–
With the distributed implementation from the acquisition to the display, the effects of better image visual quality, fast image communication and display, as well as data integrity of archived CR images in PACS are achieved.
- Design Of An Automatic Vehicle Surveillance System–
The results are presented of a study for establishing the image quality required for a license plate finder and character recognizer to perform reliably.
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-14. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.