Lei Zhu

64

Papers

438

Total Citations

1

Affiliations

Affiliations

The Chinese University of Hong Kong

Papers (64)

Bidirectional Feature Pyramid Network with Recurrent Attention Residual Modules for Shadow Detection

Meissonic: Revitalizing Masked Generative Transformers for Efficient High-Resolution Text-to-Image Synthesis

Learning Diffusion Texture Priors for Image Restoration

MindTuner: Cross-Subject Visual Decoding with Visual Fingerprint and Semantic Correction

POSTA: A Go-to Framework for Customized Artistic Poster Generation

Inverse Rendering of Glossy Objects via the Neural Plenoptic Function and Radiance Fields

Scribble Hides Class: Promoting Scribble-Based Weakly-Supervised Semantic Segmentation

V2C-CBM: Building Concept Bottlenecks with Vision-to-Concept Tokenizer

Turbo2K: Towards Ultra-Efficient and High-Quality 2K Video Synthesis

Detect Any Mirrors: Boosting Learning Reliability on Large-Scale Unlabeled Data with an Iterative Data Engine

SnowMaster: Comprehensive Real-world Image Desnowing via MLLM with Multi-Model Feedback Optimization

Residual Diffusion Deblurring Model for Single Image Defocus Deblurring

Agile Multi-Source-Free Domain Adaptation

Effective Comparative Prototype Hashing for Unsupervised Domain Adaptation

Domain-Agnostic Mutual Prompting for Unsupervised Domain Adaptation

Low-Res Leads the Way: Improving Generalization for Super-Resolution by Self-Supervised Learning

Genuine Knowledge from Practice: Diffusion Test-Time Adaptation for Video Adverse Weather Removal

Beyond Text: Frozen Large Language Models in Visual Signal Comprehension

A Non-Local Low-Rank Framework for Ultrasound Speckle Reduction

Direction-Aware Spatial Context Features for Shadow Detection

Leveraging the Invariant Side of Generative Zero-Shot Learning

Depth-Attentional Features for Single-Image Rain Removal

A Multi-Task Mean Teacher for Semi-Supervised Shadow Detection

Triple-Cooperative Video Shadow Detection

Involution: Inverting the Inherence of Convolution for Visual Recognition

Learning the Superpixel in a Non-Iterative and Lifelong Manner

Cross-Domain Gradient Discrepancy Minimization for Unsupervised Domain Adaptation

A Multi-Task Network for Joint Specular Highlight Detection and Removal

Distinguishing Unseen From Seen for Generalized Zero-Shot Learning

Weakly Supervised Object Localization As Domain Adaption

BoostMIS: Boosting Medical Image Semi-Supervised Learning With Adaptive Pseudo Labeling and Informative Active Annotation

RSCFed: Random Sampling Consensus Federated Semi-Supervised Learning

BiFormer: Vision Transformer With Bi-Level Routing Attention

Neural Preset for Color Style Transfer

SCOTCH and SODA: A Transformer Video Shadow Detection Framework

Video Dehazing via a Multi-Range Temporal Alignment Network With Physical Prior

Masked Image Training for Generalizable Deep Image Denoising

Joint Bi-Layer Optimization for Single-Image Rain Streak Removal

Leveraging Weak Semantic Relevance for Complex Video Event Classification

Saliency Pattern Detection by Ranking Structured Trees

Deep Multi-Model Fusion for Single-Image Dehazing

Unifying Nonlocal Blocks for Neural Networks

VIL-100: A New Dataset and a Baseline Model for Video Instance Lane Detection

MT-ORL: Multi-Task Occlusion Relationship Learning

Mitigating Intensity Bias in Shadow Detection via Feature Decomposition and Reweighting

Sparse Sampling Transformer with Uncertainty-Driven Ranking for Unified Removal of Raindrops and Rain Streaks

Towards High-Quality Specular Highlight Removal by Leveraging Large-Scale Synthetic Data

Snow Removal in Video: A New Dataset and A Novel Method

Video Adverse-Weather-Component Suppression Network via Weather Messenger and Adversarial Backpropagation

Bagging Regional Classification Activation Maps for Weakly Supervised Object Localization

Harmonizer: Learning to Perform White-Box Image and Video Harmonization

Rethinking Video Rain Streak Removal: A New Synthesis Model and a Deraining Network with Video Rain Prior

Interpretable Open-Set Domain Adaptation via Angular Margin Separation

RoGSplat: Learning Robust Generalizable Human Gaussian Splatting from Sparse Multi-View Images

A Simple Data Augmentation for Feature Distribution Skewed Federated Learning

Medical World Model

GenHaze: Pioneering Controllable One-Step Realistic Haze Generation for Real-World Dehazing

GlassWizard: Harvesting Diffusion Priors for Glass Surface Detection

Toward Fair and Accurate Cross-Domain Medical Image Segmentation: A VLM-Driven Active Domain Adaptation Paradigm

Towards Realistic Semi-supervised Medical Image Classification

AGLLDiff: Guiding Diffusion Models Towards Unsupervised Training-free Real-world Low-light Image Enhancement

PromptHaze: Prompting Real-world Dehazing via Depth Anything Model

Dynamic Target Distribution Estimation for Source-Free Open-Set Domain Adaptation

Learning Together Securely: Prototype-Based Federated Multi-Modal Hashing for Safe and Efficient Multi-Modal Retrieval