Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileAcademic lineage
Possible advisorsa guess from early papers, not confirmed
- Lei ZhangSuggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works29 from public data
- SeeSR: Towards Semantics-Aware Real-World Image Super-Resolution438
A degradation-aware prompt extractor is trained, which can generate accurate soft and hard semantic prompts even under strong degradation, and a semantics-aware approach is presented to better preserve the semantic fidelity of generative real-world image super-resolution.
- 122
- 117
- Polarization image demosaicking using polarization channel difference prior84
The proposed polarization demosaicking method consists of only 16 convolution operations, which makes it fast and parallelizable for GPU acceleration, and theoretically proves that this prior is physical valid.
- Adversarial Diffusion Compression for Real-World Image Super-Resolution73
A novel Real-ISR method, AdcSR, is proposed by distilling the one-step diffusion network OSEDiff into a streamlined diffusion-GAN model under the authors' Adversarial Diffusion Compression (ADC) framework, which effectively reduces complexity by 73% in inference time, 78% in computation, and 74% in parameters, while preserving the model’s generation capability.
- 40
- Polarization-guided road detection network for LWIR division-of-focal-plane camera33
Experiments demonstrate that the proposed polarization-guided road detection network outperforms state-of-the-art real-time segmentation networks with fewer parameters and faster speed.
- Fine-Structure Preserved Real-World Image Super-Resolution Via Transfer Vae Training27
A Transfer VAE Training (TVT) strategy to transfer the $8 \times$ downsampled VAE into a $4 \times$ one while adapting to the pre-trained UNet, which significantly improves fine-structure preservation, while requiring fewer FLOPs than state-of-the-art onestep diffusion models.
- NTIRE 2025 the 2nd Restore Any Image Model (RAIM) in the Wild Challenge24
This challenge established a new benchmark for real-world image restoration, featuring diverse scenarios with and without reference ground truth, and tasked participants with restoring real-captured images suffering from complex and unknown degradations.
- NTIRE 2024 Restore Any Image Model (RAIM) in the Wild Challenge22
The RAIM challenge constructed a benchmark for image restoration in the wild, including real-world images with/without reference ground truth in various scenarios from real applications, which improved the state-of-the-art restoration performance and obtained unanimous recognition from all 18 judges.
- Improving the Stability and Efficiency of Diffusion Models for Content Consistent Super-Resolution18
This work proposes to partition the generative SR process into two stages, where the DM is employed for reconstructing image structures and the GAN is employed for improving fine-grained details, and proposes a non-uniform timestep sampling strategy in the first stage.
- 17
- 10
- Progressive Rendering Distillation: Adapting Stable Diffusion for Instant Text-to-Mesh Generation without 3D Data8
Progressive Rendering Distillation (PRD) is proposed, eliminating the need for 3D ground-truths by distilling multi-view diffusion models and adapting SD into a native 3D generator, which outperforms previous text-to-3D generators in both efficiency and quality.
- 7
- VOSR: A Vision-Only Generative Model for Image Super-Resolution6
This work proposes VOSR, a Vision-Only generative framework for SR, and shows that high-quality generative SR can be achieved without multimodal pretraining.
- GDPO-SR: Group Direct Preference Optimization for One-Step Generative Image Super-Resolution4
Group Direct Preference Optimization (GDPO) is proposed, a novel approach to integrate RL into one-step generative ISR model training and an attribute-aware reward function is designed to dynamically evaluate the score of each sample based on its statistics of smooth and texture areas.
- 4
- Restoration Adaptation for Semantic Segmentation on Low Quality Images3
Restoration Adaptation for Semantic Segmentation (RASS) is proposed, which effectively integrates semantic image restoration into the segmentation process, enabling high-quality semantic segmentation on the LQ images directly.
- NSARM: Next-Scale Autoregressive Modeling for Robust Real-World Image Super-Resolution3
This work introduces a robust Real-ISR framework, namely Next-Scale Autoregressive Modeling (NSARM), and demonstrates much higher robustness to the quality of input images, showing stronger generalization performance.
- 3
- 1
- GGT-100K: Generative Ground Truth for Generalizable Real-World Image Restoration1
This work proposes Generative Ground Truth (GGT) by using generative multimodal foundation models (MFMs) to produce high-quality (HQ) targets from real-world low-quality (LQ) images and shows that GGT-100K consistently improves the real-world generalization of a wide range of IR models.
- –
- –
- –
- –
- –
- –
Publication data from OpenAlex; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-15. 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.