roytseng-tw/Detectron.pytorch
A PyTorch implementation of Facebook's Detectron for training and deploying object detection, instance segmentation, and keypoint estimation models.
Not currently ranked — collecting fresh signals.
star history
This repository provides a pure PyTorch implementation of Facebook Research’s Detectron framework, enabling object detection, instance segmentation, and human pose estimation. It supports training models from scratch using custom datasets and inference using pretrained Detectron weight files. The implementation includes support for multi-GPU training, memory-efficient batching techniques, and various ROI pooling methods aligned with the original Caffe2-based Detectron.
Frequently asked
- What is roytseng-tw/Detectron.pytorch?
- A PyTorch implementation of Facebook's Detectron for training and deploying object detection, instance segmentation, and keypoint estimation models.
- Is Detectron.pytorch open source?
- Yes — roytseng-tw/Detectron.pytorch is open source, released under the MIT license.
- What language is Detectron.pytorch written in?
- roytseng-tw/Detectron.pytorch is primarily written in Python.
- How popular is Detectron.pytorch?
- roytseng-tw/Detectron.pytorch has 2.8k stars on GitHub.
- Where can I find Detectron.pytorch?
- roytseng-tw/Detectron.pytorch is on GitHub at https://github.com/roytseng-tw/Detectron.pytorch.