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
- Suggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works44 from public data
- 186
- Neural Degradation Representation Learning for All-in-One Image Restoration98
This paper proposes to learn a neural degradation representation (NDR) that captures the underlying characteristics of various degradations that adaptively decomposes different types of degradations, similar to a neural dictionary that represents basic degradation components.
- Exposure-Consistency Representation Learning for Exposure Correction53
A proposed ECP module to consistently learn the representation of both underexposure and overexposure in the feature space and an Exposure-Consistency Constraining strategy that augments the various local region exposures and then constrains the feature representation change between the exposure augmented image and the original one are developed.
- 52
- Towards Bidirectional Arbitrary Image Rescaling: Joint Optimization and Cycle Idempotence42
The proposed method is the first to treat arbitrary rescaling, both upscaling and downscaling, as one unified process, and is able to learn up Scaling and Downscaling simultaneously and achieve bidirectional arbitrary image rescaling.
- 41
- DSANet: Dynamic Segment Aggregation Network for Video-Level Representation Learning35
The proposed DSA module is an efficient plug-and-play module that can be combined with the off-the-shelf clip-based models to perform powerful long-range modeling with minimal overhead and benefit various video recognition models significantly.
- 34
- Mutual-Guided Dynamic Network for Image Fusion33
A novel mutual-guided dynamic network (MGDN) is proposed for image fusion, which allows for effective information utilization across different locations and inputs and outperforms existing methods on four image fusion tasks.
- Spectral-depth imaging with deep learning based reconstruction30
This is the first attempt to capture 5D information (3D space + 1D spectrum +1D time) with a miniaturized apparatus and without active illumination.
- Generalized Lightness Adaptation with Channel Selective Normalization28
A novel generalized lightness adaptation algorithm that extends conventional normalization techniques through a channel filtering design, dubbed Channel Selective Normalization (CSNorm), which purposely normalizes the statistics of lightness-relevant channels and keeps other channels unchanged, so as to improve feature generalization and discrimination.
- PolarFree: Polarization-based Reflection-Free Imaging27
A large-scale dataset, PolaRGB, is constructed for Polarization-based reflection removal of RGB images, which enables us to train models that generalize effectively across a wide range of real-world scenarios and introduces PolarFree, which leverages diffusion process to generate reflection-free cues for accurate reflection removal.
- 26
- 25
- SIR-Former: Stereo Image Restoration Using Transformer25
A stereo image restoration transformer (SIR-Former) is constructed to effectively exploit the cross-view correlations and can remarkably boost quantitative and qualitative quality on various image restoration tasks, which demonstrate the effectiveness of the proposed framework.
- Stereo Video Super-Resolution via Exploiting View-Temporal Correlations24
A novel Stereo Video Super-Resolution Network (SVSRNet) is proposed to fulfill the StereoVSR task via exploiting view-temporal correlations to enforce consistency constraint of superresolved stereo videos.
- Structure- and Texture-Aware Learning for Low-Light Image Enhancement21
A Structure and Texture Aware Network (STAN) is proposed for low-light image enhancement, which consists of a structure sub- network and a texture sub-network and a dual distillation scheme that assists the learning process of two sub-networks via counterparts' normal-light structure and texture representations.
- 17
- Continuous Spectral Reconstruction from RGB Images via Implicit Neural Representation16
This paper introduces a novel continuous spectral representation, NeSR, and embraces the concept of implicit function and implements a parameterized embodiment with a neural network to extend the flexibility of spectral reconstruction by enabling an arbitrary number of spectral bands as the target output.
- 16
- Machine Learning Boosted Entropy-Engineered Synthesis of CuCo Nanometric Solid Solution Alloys for Near-100% Nitrate-to-Ammonia Selectivity15
A data-driven approach incorporating sample balancing enhancement techniques and multilayer perceptron (MLP) algorithms to improve the model's ability to handle imbalanced data, significantly boosting the efficiency of experimental parameter optimization is utilized.
- PhotoAgent: Exploratory Visual Aesthetic Planning with Large Vision Models9
This work formulates autonomous image editing as a long-horizon decision-making problem, and introduces UGC-Edit, an aesthetic evaluation benchmark consisting of 7,000 photos and a learned aesthetic reward model that consistently improves both instruction adherence and visual quality compared with baseline methods.
- 9
- Continuous Spatial-Spectral Reconstruction via Implicit Neural Representation9
The proposed neural spatial-spectral representation (NeSSR), which projects the deep features extracted from low-resolution inputs to the corresponding intensity values under target 3D coordinates, allows for spectral image reconstruction with arbitrary resolutions in both spatial and spectral dimensions for the first time.
- 8
- Beyond Surrogate Gradients: Fully Differentiable Token Pruning for Vision-Language Models7
DiffPrune is proposed, which reformulates pruning as continuous control of token information instead of discrete selection learning, and introduces an Information Throttler that modulates each token using variance-preserving noise conditioned on importance scores, where higher scores induce less information suppression during training.
- 7
- 7
- EdiTor: Edge-guided Transformer for Ghost-free High Dynamic Range Imaging7
A novel Edge-guided Transformer framework (EdiTor) customized for ghost-free HDR reconstruction, where the long-range motions across different exposures can be delicately modeled by incorporating the edge prior, which outperforms state-of-the-art methods both quantitatively and qualitatively.
- 6
- Unsupervised Diffusion-Based Degradation Modeling for Real-World Super-Resolution5
This paper proposes an unsupervised diffusion-based degradation modeling framework (UDDM) to effectively capture real-world degradation distributions and introduces a physics-based dynamic degradation module (P-DDM) that adaptively models content-aware degradation, ensuring both content and structural accuracy.
- 4
- High-Resolution and Few-Shot View Synthesis from Asymmetric Dual-Lens Inputs3
This paper proposes a consistency-aware training strategy, which integrates a dual-lens-consistent loss to regularize the 3D-GS optimization and designs a multi-reference-guided refinement module to select proper telephoto and wide-angle guided images from training samples based on the camera pose distances, and then exploit their information for high-frequency detail enhancement.
- MIPI 2023 Challenge on RGBW Fusion: Methods and Results3
This paper summarizes and reviews the RGBW Joint Fusion and Denoise track on MIPI 2023 and describes the models developed in this challenge, including PSNR, SSIM, LPIPS, and KLD.
- Transition-constant Normalization for Image Enhancement3
This work presents a novel Transition-Constant Normalization (TCN), which consists of two streams of normalization operations arranged under an invertible constraint, along with a feature sub-sampling operation that satisfies the normalization constraint.
- Computational Spectral-Depth Imaging with a Compact System2
A compact imaging system is developed to enable simultaneous acquisition of the spectral and depth information in real-time with high resolution using only two commercial cameras and relying on an efficient computational reconstruction algorithm with deep learning.
- Uni-ISP: Towards Unifying the Learning of ISPs from Multiple Mobile Cameras1
Uni-ISP is introduced, a novel pipeline that unifies ISP learning for diverse mobile cameras, delivering a highly accurate and adaptable processor and unlocking new applications previously inaccessible to conventional learned ISPs.
- –
- MotionDreamer: Generating Continuous Intermediate Motions for Complex Motion Interpolation–
MotionDreamer is a novel framework that instills effective motion priors for all intermediate frames to video generators by regularizing diffusion training with ground truth continuous intermediate motions and collects the first open-source large-scale dataset, VFIVid, for complex motion interpolation.
- –
- –
- –
- InstantRetouch: Efficient and High-Fidelity Instruction-Guided Image Retouching with Bilateral Space–
This work proposes an efficient and fidelity-preserving retouching method using bilateral space manipulation, which is both compact and content-decoupled and can avoid content drift, significantly improve latency, and generate visually pleasing edits, while maintaining a high level of fidelity.
- High-resolution and real-time spectral-depth imaging with a compact system–
This is the first attempt to capture 5D information with a miniaturized apparatus and without active illumination, and reliance on an elaborate computational reconstruction algorithm with deep learning results are accurate and reliable.
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.