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
- Xiaoou TangSuggested from co-authorship
- William T. FreemanSuggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works105 from public data
- Video Enhancement with Task-Oriented Flow1,801
T task-oriented flow (TOFlow), a motion representation learned in a self-supervised, task-specific manner, is proposed, which outperforms traditional optical flow on standard benchmarks as well as the Vimeo-90K dataset in three video processing tasks.
- 1,394
- Unprocessing Images for Learned Raw Denoising572
This work presents a technique to “unprocess” images by inverting each step of an image processing pipeline, thereby allowing us to synthesize realistic raw sensor measurements from commonly available Internet photos.
- Pix3D: Dataset and Methods for Single-Image 3D Shape Modeling529
A novel model is designed that simultaneously performs 3D reconstruction and pose estimation; this multi-task learning approach achieves state-of-the-art performance on both tasks.
- 320
- Single Image 3D Interpreter Network292
This work proposes 3D INterpreter Network (3D-INN), an end-to-end framework which sequentially estimates 2D keypoint heatmaps and 3D object structure, trained on both real 2D-annotated images and synthetic 3D data, and achieves state-of-the-art performance on both 2DKeypoint estimation and3D structure recovery.
- A computational approach for obstruction-free photography253
A unified computational approach for taking photos through reflecting or occluding elements such as windows and fences by instructing the user to take a short image sequence while slightly moving the camera.
- 244
- EmbodiedScan: A Holistic Multi-Modal 3D Perception Suite Towards Embodied AI206
This work introduces EmbodiedScan, a multi-modal, ego-centric 3D perception dataset and benchmark for holistic 3D scene understanding, and a baseline framework named Embodied Perceptron, capable of processing an arbitrary number of multi-modal inputs and demonstrating remarkable 3D perception capabilities.
- Teaching Large Language Models to Regress Accurate Image Quality Scores Using Score Distribution183
The distribution-based Depicted image Quality Assessment model for Score regression (DeQA-Score) is developed, showing that DeQA-Score stably outperforms baselines in score regression and can predict the score distribution that closely aligns with human annotations.
- How to Train Neural Networks for Flare Removal113
Experiments show that the data synthesis approach is critical for accurate flare removal, and that models trained with the technique generalize well to real lens flares across different scenes, lighting conditions, and cameras.
- 99
- Best-Buddies Similarity—Robust Template Matching Using Mutual Nearest Neighbors89
The Best-Buddies Similarity (BBS) is proposed, a useful, robust, and parameter-free similarity measure between two sets of points that is robust against complex geometric deformations and high levels of outliers.
- 88
- CineMaster: A 3D-Aware and Controllable Framework for Cinematic Text-to-Video Generation81
To overcome the scarcity of in-the-wild datasets with 3D object motion and camera pose annotations, this work carefully establish an automated data annotation pipeline that extracts 3D bounding boxes and camera trajectories from large-scale video data.
- Joint Bilateral Learning for Real-Time Universal Photorealistic Style Transfer64
Compared to the state of the art, the proposed end-to-end model for photorealistic style transfer produces visually superior results and is three orders of magnitude faster, enabling real-time performance at 4K on a mobile phone.
- 55
- Refraction Wiggles for Measuring Fluid Depth and Velocity from Video54
Different from previous approaches for measuring refractive flow, these methods operate directly on videos captured with ordinary cameras, do not require auxiliary sensors, light sources or designed backgrounds, and can correctly detect the motion and location of refractive fluids even when they are invisible to the naked eye.
- AligNeRF: High-Fidelity Neural Radiance Fields via Alignment-Aware Training53
This work proposes the corresponding solutions to marry the multilayer perceptron (MLP) with convolutional layers which can encode more neighborhood information while reducing the total number of parameters, and proposes a novel training strategy to address misalignment caused by moving objects or small camera calibration errors.
- CamCloneMaster: Enabling Reference-based Camera Control for Video Generation51
This work proposes CamCloneMaster, a framework that enables users to replicate camera movements from reference videos without requiring camera parameters or test-time fine-tuning, and presents the Camera Clone Dataset, a large-scale synthetic dataset designed for camera clone learning, encompassing diverse scenes, subjects, and camera movements.
- 50
- Defocus Map Estimation and Deblurring from a Single Dual-Pixel Image50
This work presents a method that takes as input a single dual-pixel image, and simultaneously estimates the image’s defocus map—the amount of defocus blur at each pixel—and recovers an all-in-focus image, which improves upon prior works for bothdefocus map estimation and blur removal, despite being entirely unsupervised.
- Enhancing Descriptive Image Quality Assessment With a Large-Scale Multi-Modal Dataset45
Experimental results demonstrate that EDQA significantly outperforms traditional score-based methods, prior VLM-based IQA models, and proprietary GPT-4V in distortion identification, instant rating, and reasoning tasks.
- HDRFlow: Real-Time HDR Video Reconstruction with Large Motions43
HDRFlow is the first real-time HDR video reconstruction method for video sequences captured with alternating exposures, capable of processing 720p resolution inputs at 25ms, and outperforms previous methods on standard benchmarks.
- Visual Dynamics: Stochastic Future Generation via Layered Cross Convolutional Networks37
The probabilistic model makes it possible for us to sample and synthesize many possible future frames from a single input image, and to synthesize realistic movement of objects, a novel network structure is proposed, namely a Cross Convolutional Network.
- 37
- Real-time Localized Photorealistic Video Style Transfer35
This work presents a novel algorithm, based on a deep neural network architecture inspired by recent work in photorealistic style transfer, that works on arbitrary inputs without runtime optimization once trained on a diverse dataset of artistic styles.
- Sparse representation over discriminative dictionary for stereo matching35
The proposed matching cost outperforms traditional matching costs, the discriminative dictionary learning model is more suitable than previous dictionary learning models for stereo matching, and the proposed stereo method ranks the third place on the Middlebury benchmark v3 in quarter resolution up to the submitting, and achieves the best accuracy on 30 classic stereo images.
- 32
- 29
- 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
- 24
- 23
- 22
- 22
- 22
- Fast and High Quality Image Denoising via Malleable Convolution20
This work presents Malleable Convolution (MalleConv), which performs spatial-varying processing with minimal computational overhead, and uses it to build an efficient denoising network the authors call MalleNet, which achieves high-quality results without very deep architectures.
- 19
- 17
- 16
- 15
- 13
- 12
- 12
- 12
- From Sim-to-Real: Toward General Event-based Low-light Frame Interpolation with Per-scene Optimization11
This work proposes a novel per-scene optimization strategy that utilizes the internal statistics of a sequence to handle degraded event data under low-light conditions, improving the generalizability to different lighting and camera settings.
- 11
- Learning-based lens wavefront aberration recovery10
A lightweight and practical learning-based method to recover the wavefront aberration for an imaging system from a single intensity measurement, named LWNet, which works well for wavefront aberration estimation of real imaging systems and consistently outperforms prior learning-based methods.
- 10
- Reconstruct-and-Generate Diffusion Model for Detail-Preserving Image Denoising9
This work proposes a novel approach called the Reconstruct-and-Generate Diffusion Model (RnG), which leverages a reconstructive denoising network to recover the majority of the underlying clean signal and employs a diffusion algorithm to generate residual high-frequency details, thereby enhancing visual quality.
- A Physics-Informed Blur Learning Framework for Imaging Systems8
This paper introduces a novel wavefront-based PSF model accompanied by an optimization strategy, both reducing optimization complexity and improving estimation accuracy, and demonstrates improvements in image quality in simulation and also showcases noticeable visual quality improvements on real captured images.
- 8
- 8
- 7
- 7
- 7
- 5
- 5
- 4
- 4
- An example-based approach to 3D man-made object reconstruction from line drawings4
An example-based approach to reconstructing 3D object, either planar or curved, from a single-view line drawing with sketch errors, which can handle a wide range of 3D man-made objects including curved components.
- AdaptiveAE: An Adaptive Exposure Strategy for HDR Capturing in Dynamic Scenes3
AdaptiveAE is a reinforcement learning-based method that optimizes the selection of shutter speed and ISO combinations to maximize HDR reconstruction quality in dynamic environments and can adaptively select optimal ISO and shutter speed sequences based on a user-defined exposure time budget.
- 3
- An End-to-End Optimized Lensless System for Privacy-Preserving Face Verification2
This paper proposes a face center alignment scheme, an augmentation curriculum to build robustness against variations, and a knowledge distillation method to smooth optimization and enhance performance that outperforms two-stage lensless verification while enhancing privacy and efficiency.
- 2
- 2
- 2
- LenslessFace: An End-to-End Optimized Lensless System for Privacy-Preserving Face Verification2
This paper proposes a face center alignment scheme, an augmentation curriculum to build robustness against variations, and a knowledge distillation method to smooth optimization and enhance performance for privacy-preserving face verification directly on encoded lensless captures.
- 1
- 1
- 1
- 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.
- 1
- 1
- 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.
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- Revisiting the generalization problem of low-level vision models through the lens of image deraining–
- –
- –
- –
- –
- –
- –
- InstantRetouch: Efficient and High-Fidelity Instruction-Guided Image Retouching with Bilateral Space–
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
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 2026-10-13. 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.