Spatiotemporal modeling of spatial transcriptomics

The code for reconstruction algorithm and post-processing of electron tomographic series of colloidal particles acquired in liquid

Photogrammetric Computer Vision Framework

A collaboration friendly studio for NeRFs

🎓Automatically Update CV Papers Daily using Github Actions (Update Every 12th hours)

Cybervision can generate a 3D model from two photos of an object

A collection of some useful (or maybe just interesting) computer vision tools.

[CVPR 2024] RoMa: Robust Dense Feature Matching; RoMa is the robust dense feature matcher capable of estimating pixel-dense warps and reliable certainties for almost any image pair.

3D Reconstruction Software

MagellanMapper is a graphical interface for 3D bioimage annotation, atlas registration, and regional quantification

Open source Structure-from-Motion pipeline

A Modular Framework for 3D Gaussian Splatting and Beyond

3D Shape Completion - adaptation and improvement of DiffComplete model.

A procedural Blender pipeline for photorealistic training image generation

Awesome papers for markerless animal motion capture and 3D reconstruction.

Code of [CVPR 2024] "Animatable Gaussians: Learning Pose-dependent Gaussian Maps for High-fidelity Human Avatar Modeling"

Implantation of stereo camera calibration and 3D reconstruction

Gaussian Opacity Fields: Efficient and Compact Surface Reconstruction in Unbounded Scenes

RGBD-3DGS-SLAM is a monocular SLAM system leveraging 3D Gaussian Splatting (3DGS) for accurate point cloud and visual odometry estimation. By integrating neural networks, it estimates depth and camera intrinsics from RGB images alone, with optional support for additional camera information and depth maps.

Helical Indexing using Layer Lines