Paper 'Understanding Masked Autoencoders From a Local Contrastive Perspective' published.
Paper 'OV-PARTS: Towards Open-Vocabulary Part Segmentation' accepted to NeurIPS Dataset and Benchmark Track 2023.
Paper 'Patch-based separable transformer for visual recognition' published in T-PAMI.
Two papers accepted to ICCV 2021: 'Vision Transformer with Progressive Sampling' and 'Aggregation with Feature Detection'.
Co-developed and open-sourced MMOCR, a comprehensive toolbox for text detection, recognition, and understanding; paper published at ACM MM 2021 (Open Source Competition Track).
Paper 'RobustScanner: Dynamically Enhancing Positional Clues for Robust Text Recognition' accepted to ECCV 2020.
Paper 'HOSE-Net: Higher Order Structure Embedded Network for Scene Graph Generation' published at ACM MM 2020.
Paper 'Geometry Normalization Networks for Accurate Scene Text Detection' accepted to ICCV 2019.