Current frame
Computer vision and multimodal learning researcher
Atlas / People / Detail
Jihao Liu
Public sources identify Jihao Liu as a researcher working on computer vision, multimodal learning, and vision-language topics. A public GitHub profile under `jihaonew` lists "CUHK, MMLab" and a personal website, and DBLP shows publications from 2020 to 2024 across image modeling, vision pretraining, and multimodal alignment.
Profile status: updated
Extended note
Jihao Liu is publicly identifiable through a DBLP author profile and a GitHub profile under `jihaonew` labeled "Jihao Liu." DBLP credits him on research spanning computer vision and multimodal learning, including Rotate-and-Render (CVPR 2020), UniNet and TokenMix (ECCV 2022), MixMAE (CVPR 2023), GeoMIM (ICCV 2023), GLID (CVPR 2024), and MM-Instruct (2024). He is also listed as a coauthor of the Seed1.5-VL Technical Report (2025). The GitHub profile lists "CUHK, MMLab" in Hong Kong and links to a personal website; because that affiliation is self-declared rather than from an official institutional page, it is used here conservatively for bio context only.