2023.09.04 Codes for Expedit-Mask2Former, Expedit-Video-Swin-Transformer and Expedit-OneFormer are released.
2023.04.14 Demo for Expedit-SAM is released! Have fun!🍺🍺
2023.04.13 Code for Expedit-SAM is released. We elegantly expedite Segment Anything Model (SAM) by using our method without any training.
2023.03.24 Code for Expedit-DINO is released.
2023.03.21 Codes for Expedit-Segmenter, Expedit-SWAG, and Expedit-DPT are released.
2022.10.03 "Expediting Large-Scale Vision Transformer for Dense Prediction without Fine-tuning" has been released to arXiv.
2022.09.15 "Expediting Large-Scale Vision Transformer for Dense Prediction without Fine-tuning" has been accepted at NeurIPS 2022. 🎉🎉🎉
If you find this work useful in your research, please consider citing:
@article{liang2022expediting,
title={Expediting large-scale vision transformer for dense prediction without fine-tuning},
author={Liang, Weicong and Yuan, Yuhui and Ding, Henghui and Luo, Xiao and Lin, Weihong and Jia, Ding and Zhang, Zheng and Zhang, Chao and Hu, Han},
journal={arXiv preprint arXiv:2210.01035},
year={2022}
}