tjiiv-cprg/EPro-PnP
EPro-PnP is a PyTorch implementation of a differentiable probabilistic Perspective-n-Points layer for monocular 6DoF object pose estimation.
Not currently ranked — collecting fresh signals.
star history
EPro-PnP provides a continuous counterpart to the categorical Softmax layer for 6DoF pose estimation networks. It formulates perspective-n-points as an end-to-end learnable layer using Gauss-Newton and Levenberg-Marquardt optimization within the network, enabling gradient backpropagation through the geometric solver. The method achieved Best Student Paper at CVPR 2022 and is designed for monocular object pose estimation tasks.
Frequently asked
- What is tjiiv-cprg/EPro-PnP?
- EPro-PnP is a PyTorch implementation of a differentiable probabilistic Perspective-n-Points layer for monocular 6DoF object pose estimation.
- Is EPro-PnP open source?
- Yes — tjiiv-cprg/EPro-PnP is open source, released under the Apache-2.0 license.
- What language is EPro-PnP written in?
- tjiiv-cprg/EPro-PnP is primarily written in Python.
- How popular is EPro-PnP?
- tjiiv-cprg/EPro-PnP has 1.2k stars on GitHub.
- Where can I find EPro-PnP?
- tjiiv-cprg/EPro-PnP is on GitHub at https://github.com/tjiiv-cprg/EPro-PnP.