CVPR 2026 Highlights Guangdong as He Kaiming and GDUT Team Stand Out
Original: 今年CVPR看点是广东：何恺明再获至高大奖，广工大打破大厂名校垄断
CVPR 2026 awards spotlight 4D reconstruction, 3D generation, ResNet, YOLO, and Chinese researchers.
CVPR 2026 named Google DeepMind’s D4RT as Best Paper for fast dynamic 4D scene reconstruction from video. Honorable mentions included Meta’s SAM 3D and NVIDIA’s NitroGen, while TRELLIS.2 won Best Student Paper. The article emphasizes Chinese researcher visibility, ResNet and YOLO receiving the Longuet-Higgins Prize, and a GDUT-led undergraduate-heavy ChordEdit team breaking through among major labs and elite universities.
This report compiles the major awards of CVPR 2026 and several noteworthy trends. The Best Paper award went to D4RT, a dynamic 4D scene reconstruction model completed by Google DeepMind in collaboration with UCL, Oxford, and others, aiming to reconstruct three-dimensional geometry, motion, point clouds, camera parameters, and point trajectories that change over time from ordinary video. Its key design is to first encode an entire video into a global scene representation, then use a lightweight decoder to answer on demand the 3D position of a specific point in time and space, avoiding the need to build separate decoders for different subtasks. The report states that it can perform pose estimation at over 200 FPS on an A100, noticeably faster than VGGT and MegaSaM, and refreshes SOTA on dynamic 4D reconstruction and tracking tasks.
Free shows the 3-line summary; Pro unlocks the full deep summary (~300 words) so you never have to click through.
See Pro plans →Want the original English / full article?
Read on 量子位 QbitAI →Related
Summaries are AI-generated; the original article is authoritative.