Scholar
Xiang An
Google Scholar ID: 1ckaPgwAAAAJ
DeepGlint
Citations & Impact
Citations
891
H-index
13
i10-index
13
Publications
20
Co-authors
14
list available
Publications
Resume
Research Experience
- Contributed to the development of LLaVA-OneVision-1.5, a fully open framework for democratized multimodal training
- Enhanced OCR and localization capabilities of Large Pretrained Vision Transformers (ViT) for ICCV2025 with the RICE method
- Trained large-scale vision models using weak supervision on LAION400M and COYO700M for ECCV2024 with the MLCD approach
- Proposed UNICOM: Universal and Compact Representation Learning for Image Retrieval, presented at ICLR2023
- Developed Partial FC: an efficient and robust distributed hybrid parallel algorithm for large-scale face recognition trained on WebFace260M
