czczup/ViT-Adapter
A Vision Transformer adapter module designed to enhance dense prediction tasks including object detection and semantic segmentation.
Not currently ranked — collecting fresh signals.
star history
This project implements an adapter architecture that enhances standard Vision Transformers for dense prediction tasks such as object detection, semantic segmentation, and instance segmentation. Published at ICLR 2023, it achieves state-of-the-art results on benchmarks like COCO and ADE20K by introducing additional adapter modules that improve feature representation for pixel-level predictions.
Frequently asked
- What is czczup/ViT-Adapter?
- A Vision Transformer adapter module designed to enhance dense prediction tasks including object detection and semantic segmentation.
- Is ViT-Adapter open source?
- Yes — czczup/ViT-Adapter is an open-source project tracked on heatdrop.
- What language is ViT-Adapter written in?
- czczup/ViT-Adapter is primarily written in Python.
- How popular is ViT-Adapter?
- czczup/ViT-Adapter has 1.5k stars on GitHub.
- Where can I find ViT-Adapter?
- czczup/ViT-Adapter is on GitHub at https://github.com/czczup/ViT-Adapter.