My blog

Fork of NVIDIA Isaac GR00T N1.6 (paper). Replaces the flow matching (velocity MSE) training loss in GR00T's action head with drifting loss, enabling single-step inference (no iterative ODE solver).…

PyTorch Extension Library of Optimized Graph Cluster Algorithms

OBEYED-VLA: Clutter-Resistant Vision-Language-Action Models through Object-Centric and Geometry Grounding

[TMLR] DuFal: Dual-Frequency-Aware Learning for High-Fidelity Extremely Sparse-view CBCT Reconstruction

A Curated List of Awesome Works in World Modeling, Aiming to Serve as a One-stop Resource for Researchers, Practitioners, and Enthusiasts Interested in World Modeling.

Awesome curated collection of images and prompts generated by gemini-2.5-flash-image (aka Nano Banana) state-of-the-art image generation and editing model. Explore AI generated visuals created with…

Pytorch implementation for MeanFlow

[CVPR 2025] h-Edit: Effective and Flexible Diffusion-Based Editing via Doob’s h-Transform

A simple UI for SAM 2. Give an input path to a directory of video frames, and the script will let you look through the frames, plot points, then feed the points into SAM 2.

Generate videos that interpolate between two given images

A summary of related works about flow matching, stochastic interpolants

[ISBI 2024] An implementation of SAM3D which adapts Segment Anything Model for Volumetric Medical Image Segmentation

[WACV 2024] Decoding Radiologists’ Intense Focus for Accurate CXR Diagnoses: A Controllable & Interpretable AI System

[Remote Sensing] AerialFormer: Multi-resolution Transformer for Aerial Image Segmentation

[ICRA 2024 Oral] Open-Fusion: Real-time Open-Vocabulary 3D Mapping and Queryable Scene Representation

Refine high-quality datasets and visual AI models

[CVPR 2023] This is the official PyTorch implementation for "Dynamic Focus-aware Positional Queries for Semantic Segmentation".

Official PyTorch implementation of ODISE: Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion Models [CVPR 2023 Highlight]

CEDNet: A Cascade Encoder-Decoder Network for Dense Prediction (Pattern Recognition 2024)

AIOZ-GDANCE: a large-scale dataset & baseline for music-driven group dance generation. (CVPR 2023)

Detectron2 is a platform for object detection, segmentation and other visual recognition tasks.