Hongjian Zhan

one-shot learning (1) contrastive learning (1) multi-task learning (1) few-shot learning (1) object detection (1) language modeling (1) multimodal learning (1) multi-modal learning (1) open-set recognition (1) character recognition (1) scene text recognition (1) character representation (1) glyph embedding (1) text recognition (1) visual feature extraction (1) chinese text recognition (1) visual archetype (1) vision-language reasoning (1) glyph representation (1) balanced concatenation (1)