-
Korea Advanced Institute of Science & Technology (KAIST)
- Seoul
- in/soow0n
Stars
Official PyTorch Implementation of "Better Source, Better Flow: Learning Condition-Dependent Source Distribution for Flow Matching"
[NeurIPS 2026] SpatialClaw: Rethinking Action Interface for Agentic Spatial Reasoning
Official implementation of "GARD: Geometry-Aware Representation Denoising for Robust Multi-view 3D Reconstruction"
Official implementation of "Repurposing Geometric Foundation Models for Multi-view Diffusion"
TacImag: Imagining the Sense of Touch for Robotic Manipulation
InternRobotics' open platform for building generalized navigation foundation models.
Official implementation of "Geometric Action Model for Robot Policy Learning"
Official implementation of "WorldKV: Efficient World Memory with World Retrieval and Compression"
Official code implementation for TrackCraft3R: Repurposing Video Diffusion Transformers for Dense 3D Tracking
[TMLR 2026] Survey: https://arxiv.org/pdf/2507.20198
[ICCV 25] VMBench: A Benchmark for Perception-Aligned Video Motion Generation
AllTracker is a model for tracking all pixels in a video.
[CVPR 2026] Official implementation of "VideoMaMa: Mask-Guided Video Matting via Generative Prior"
Enjoy the magic of Diffusion models!
Code for "EgoX: Egocentric Video Generation from a Single Exocentric Video"
[ICCV 2025] SpatialTrackerV2: 3D Point Tracking Made Easy
Official implementation of "Repurposing Video Diffusion Transformers for Robust Point Tracking"
[NeurIPS 2025] Official Implementation of DINO-Foresight: Looking into the Future with DINO
Official implementation of "V-Warper: Appearance-Consistent Video Diffusion Personalization via Value Warping"
[ICML 2026] Official implementation of "Deep Forcing: Training-Free Long Video Generation with Deep Sink and Participative Compression"
Official implementation of "MV-TAP: Tracking Any Point in Multi-View Videos"
Official implementation of "CAMEO: Correspondence-Attention Alignment for Multi-View Diffusion Models"
[CVPR'26] Official implementation of "Emergent Outlier View Rejection in Visual Geometry Grounded Transformers"
Official implementation of "Unified Diffusion Transformer for High-Fidelity Text-Aware Image Restoration"
[CVPR'26] Official implementation of "C3G: Learning Compact 3D Representations with 2K Gaussians"
Reference PyTorch implementation and models for DINOv3
[ICLR'26] Official implementation of "3D Scene Prompting for Scene-Consistent Camera-Controllable Video Generation"