CLIP as RNN: Segment Countless Visual Concepts without Training Endeavor (CaR)

This repo holds the implementation code of the paper CLIP as RNN: Segment Countless Visual Concepts without Training Endeavor (CaR) by Shuyang Sun, Runjia Li, Philip Torr, Xiuye Gu, and Siyang Li:

@article{clip_as_rnn,
      title={CLIP as RNN: Segment Countless Visual Concepts without Training Endeavor},
      author={Shuyang Sun and Runjia Li and Philip Torr and Xiuye Gu and Siyang Li},
      year={2023},
      eprint={2312.07661},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}