Scholar

Junting Pan (潘俊廷）

Google Scholar ID: 8Xt3TnAAAAAJ

Multimedia Lab (MMLab), The Chinese University of Hong Kong

Computer VisionMachine Learning

Citations & Impact

All-time

Citations

5,396

H-index

i10-index

Publications

Co-authors

list available

Contact

Publications

20 items

Browse publications on Google Scholar (top-right) ↗

Resume (English only)

Academic Achievements

2024: Paper 'Math-Vision' accepted to NeurIPS 2024 Datasets and Benchmarks Track; released Math-Vision benchmark for evaluating mathematical reasoning in LMMs
2024: Co-released SAM 2, a unified model for real-time, promptable video object segmentation
2023: Paper 'JourneyDB' accepted to NeurIPS 2023 Datasets and Benchmarks Track; released JourneyDB, a large-scale benchmark for multimodal generative image understanding
2022: Paper 'ST-Adapter' accepted to NeurIPS 2022, proposing efficient image-to-video transfer learning
2022: Paper 'EdgeViTs' accepted to ECCV, introducing lightweight Vision Transformers competitive with CNNs on mobile devices
2023: Paper 'Retrieving-to-Answer' accepted to ICCVW, achieving SOTA in zero-shot video question answering
Published multiple papers at top-tier conferences including NeurIPS, CVPR, ICCV, and ECCV

Co-authors

14 total