CHSAM: Efficient Scene Text Segmentation via SAM with Convolutional Adapters and Hierarchical Decoding.

Jing-Yao Zhang, Heng Zhang, Fei Yin

Browse the full ICDAR paper archive.