TOOD: Task-aligned One-stage Object Detection, ICCV2021 Oral
Official PyTorch implementation of MX-Font (Multiple Heads are Better th...
A collection of differentiable SVD methods and ICCV21 "Why Approximate M...
Exploring Classification Equilibrium in Long-Tailed Object Detection, IC...
Toward Realistic Single-View 3D Object Reconstruction with Unsupervised ...