ViTAE-Transformer/ViTDet
PyTorch implementation of Vision Transformer backbones for object detection trained with Mask RCNN on COCO.
Not currently ranked — collecting fresh signals.
star history
This repository provides an unofficial implementation of the ViTDet paper, exploring plain Vision Transformer backbones for object detection. It supports training ViT and ViTAE models with Mask RCNN on COCO dataset, reporting box and mask mAP metrics. The implementation includes configurations for multiple model sizes (Base, Small) and supports GPU and TPU training.
Frequently asked
- What is ViTAE-Transformer/ViTDet?
- PyTorch implementation of Vision Transformer backbones for object detection trained with Mask RCNN on COCO.
- Is ViTDet open source?
- Yes — ViTAE-Transformer/ViTDet is open source, released under the Apache-2.0 license.
- What language is ViTDet written in?
- ViTAE-Transformer/ViTDet is primarily written in Python.
- How popular is ViTDet?
- ViTAE-Transformer/ViTDet has 586 stars on GitHub.
- Where can I find ViTDet?
- ViTAE-Transformer/ViTDet is on GitHub at https://github.com/ViTAE-Transformer/ViTDet.