StarVector is a foundation model for SVG generation that transforms vectorization into a code generation task. Using a vision-language modeling architecture, StarVector processes both visual and te…

A sparse attention kernel supporting mix sparse patterns（block_sparse_attn-0.0.1+cu124torch2.6cxx11abiTRUE-cp311-cp311-linux_x86_64.whlアップロード用のフォーク、Releaseからダウンロード可）

The repository provides code for running inference and finetuning with the Meta Segment Anything Model 3 (SAM 3), links for downloading the trained model checkpoints, and example notebooks that sho…

TreeFormer: Single-view Plant Skeleton Estimation via Tree-constrained Graph Generation

[CVPR 2025] MINIMA: Modality Invariant Image Matching

Training library for local feature detection and matching

Official implementation for paper "Matching Ambiguity-Resilient Multi-View Stereo via Adaptive Patch Deformation"

[CVPR'24 Highlight & Best Demo Award] Gaussian Splatting SLAM

Official code for paper "InstantSfM: Fully Sparse and Parallel Structure-from-Motion"

Visual Odometry for RGBD camera. Leveraged ORB features for tracking, P3P with RANSAC for initial pose estimation, pose only BA and local map BA for optimization.

Python transparent bindings for LSD (Line Segment Detector)

Light-weight library to perform homography estimation with RANSAC from point, line or point-line correspondences

AllTracker is a model for tracking all pixels in a video.

[CVPR'23] Video Dehazing via a Multi-Range Temporal Alignment Network with Physical Prior

Repository in conjunction with "A Pytorch-Enabled Tool for Synthetic Event Camera Data Generation and Algorithm Development" for SPIE DCS 2025