search

kevin-ssy / CLIP_as_RNN

Official Implementation for CVPR 2024 paper: CLIP as RNN: Segment Countless Visual Concepts without Training Endeavor
95 stars 3 forks source link

Dataset preparation #4

Closed ZhihuaLiuEd closed 4 months ago

ZhihuaLiuEd commented 4 months ago

Hi there,

Thanks for your amazing work on CaR and congratulation on CVPR acceptance! Will you provide the dataset preparation guidance from zero-shot segmentation to referring image segmentation and referring video segmentation?

Best, Zhihua

kevin-ssy commented 4 months ago

Hi,

Thank you for your interests! You can refer to https://github.com/kevin-ssy/CLIP_as_RNN/blob/main/DATA.md for data preparation.