Official repository for the paper "MVP4D: Multi-View Portrait Video Diffusion for Animatable 4D Avatars"

Project Lyra: Open Generative 3D World Models

Scene-Centric Unsupervised Panoptic Segmentation (CVPR 2025 Highlight)

AC3D: Analyzing and Improving 3D Camera Control in Video Diffusion Transformers

This is the official implementation of SG-I2V: Self-Guided Trajectory Control in Image-to-Video Generation.

GitHub repository for our project "Efficient Degradation-aware Any Image Restoration"

[NeurIPS 2023, Spotlight] Official code release for Transient Neural Radiance Fields for Lidar View Synthesis and 3D Reconstruction.

Boosting Unsupervised Semantic Segmentation with Principal Mask Proposals (TMLR 2024)

[ECCV 2024, Oral] Official code release for Flying With Photons: Rendering Novel Views of Propagating Light.

[ICML 2024] See More Details: Efficient Image Super-Resolution by Experts Mining

CVPR NTIRE 2023 Challenge on Real-Time Super-Resolution

Semantic Self-adaptation: Enhancing Generalization with a Single Sample (TMLR 2023)

TU Darmstadt - Deep Learning: Architectures & Methods Project SS21

[NeurIPS] TTT++: When Does Self-supervised Test-time Training Fail or Thrive?

[CVPR] Towards Robust and Adaptive Motion Forecasting: A Causal Representation Perspective