Notes
I'm actively seeking a 2026 summer internship opportunity. Let's Chat!
Feel free to email me to schedule a meeting. I'm happy to chat about:
News
2025.06
Parametric Shadow Control accepted to ICCV'25, demonstrating parametric and intuitive shadow control for T2I.
2025.06
Named Outstanding Reviewer for CVPR 2025.
2025.05
Joined Adobe NextCam team (led by Marc Levoy) as summer intern. Mentored by Shumian and Zhoutong.
2025.02
Two papers accepted to CVPR 2025. Congratulations to Tianfu, Mingyang, and Jingxi.
Publications (* Indicates Equal Contribution) [Google Scholar]
Bridging GenAI and Computational Imaging & Photography (2024-Now)
Vid2Bokeh: Synthesizing Large-Scale Light Field Datasets from Videos Enables Joint Focus and Depth-of-Field Image Editing
website /
A large image bokeh dataset from video for joint focus and DoF editing on any image
Parametric Shadow Control for Portrait Generation in Text-to-Image Diffusion Models
Add a shadow knob to AI portraits without heavy compute
Flash-Split: 2D Reflection Removal with Flash Cues and Latent Separation
PDF/
Flash cues + diffusion = clean separation of reflection and transmission
Repurposing Pre-trained Video Diffusion Models for Event-based Video Interpolation
Turn event streams into smooth videos by using video diffusion model Seeing & Tracking & Reconstruction Under Challenging Imaging Condition (2023-Now)
Temporally Consistent Atmospheric Turbulence Mitigation with Neural Representations
PDF/
A neural video representation that low-passes temporal dimention to remove turbulence
Flash-Splat: 3D Reflection Removal with Flash Cues and Gaussian Splats
Use flash-induced cues to separate reflections in 3D
CodedEvents: Optimal Point-Spread-Function Engineering for 3D-Tracking with Event Cameras
Engineer the PSF so event cameras encode depth for free
Snow Removal in Video: A New Dataset and A Novel Method
A real snow video dataset and a model that removes snow consistently across frames Image Metric Design Inspired by Human Visual System (2020-Now)
Assessor360: Multi-sequence Network for Blind Omnidirectional Image Quality Assessment
A blind 360° IQA model built on multi-sequence distortion modeling
Pipal: a large-scale image quality assessment dataset for perceptual image restoration
A large Image-Quality-Assessment dataset that captures perceptual artifacts from modern image restoration methods Controllable & Efficient Image Restoration (2020 - 2022)
Super-resolution by predicting offsets: An ultra-efficient super-resolution network for rasterized images
Do super-resolution by predicting where to sample, not what to generate
Efficient image super-resolution using vast-receptive-field attention
Vast receptive field and more attention for super-resolution
Blueprint separable residual network for efficient image super-resolution
Winner of Efficient Image Super-Resolution Track @ NTIRE workshop of CVPR’23
Toward interactive modulation for photo-realistic image restoration
2D controllable deblurring and denoising with a GAN-based neural network Meet our Chief Vibe Officer @ Maryland Headquarter
|