Code for the Nerf Navigation Paper. Implements a trajectory optimiser and state estimator which use NeRFs as an environment representation

Python script for rendering 3D human poses using Blender

Code for "CosyPose: Consistent multi-view multi-object 6D pose estimation", ECCV 2020.

[PyTorch] Official implementation of CVPR2022 paper "TransFusion: Robust LiDAR-Camera Fusion for 3D Object Detection with Transformers". https://arxiv.org/abs/2203.11496

Spatio Temporal Affinity Fields (CVPR 2019)

Autolabel is a simple GUI tool to easily assign and verify *action labels* to individual frames of your self driving dataset.

Object detection achieving 44.3 mAP / 45 fps on COCO dataset

Additive margin softmax loss in pytorch

Monocular depth estimation from a single image