Embodied AI Glossary中文CVPR / ICCV / ECCV
CVPR / ICCV / ECCV（计算机视觉三大会）CommonThe three most influential international academic conferences in computer vision.
CVPR is held every year; ICCV is held in odd-numbered years; and ECCV is held in even-numbered years, originally always in Europe. All three are top venues in vision and are CCF-A tier. Embodied AI leans heavily on vision: work on 3D reconstruction, depth estimation, segmentation, video generation, and world models is heavily published at these venues, and increasingly more VLA and embodied-navigation papers are submitted to them as well, with CVPR often running dedicated embodied-AI workshops.
Example3D vision foundation models such as DUSt3R and VGGT were both published at CVPR and later widely adopted for robot perception.
- Also called
- The Big Three of Computer Vision
- Related
- Top-tier Conferences & Journals (CCF-A Venues) · NeurIPS / ICML / ICLR · Conference on Robot Learning · Computer Vision · arXiv Preprint
- Sources
- Conference on Computer Vision and Pattern Recognition - Wikipedia
CVPR official site