CVPR2023公布了12篇最佳论文候选文章。(直接点击标题可以查看原文~)
-
Ego-Body Pose Estimation via Ego-Head Pose Estimation
单位:Stanford-----------关键词:姿态估计 -
3D Registration With Maximal Cliques
单位:西北工业大学(China) -
OmniObject3D: Large-Vocabulary 3D Object Dataset for Realistic Perception, Reconstruction and Generation
单位:商汤+港中文+上海AI Lab(China)-----------关键词:三维重建 -
MobileNeRF: Exploiting the Polygon Rasterization Pipeline for Efficient Neural Field Rendering on Mobile Architectures
单位:Google-----------关键词:三维重建 -
DynIBaR: Neural Dynamic Image-Based Rendering
单位:Google -
Planning-Oriented Autonomous Driving
单位:商汤+上海AI Lab+武大(China)-----------关键词:自动驾驶 -
DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation
单位:Google -----------关键词:AIGC -
On Distillation of Guided Diffusion Models
单位:Stanford-----------关键词:AIGC -
Visual Programming: Compositional Visual Reasoning Without Training
单位:艾伦AI研究所 -
What Can Human Sketches Do for Object Detection?
单位:Surrey大学-----------关键词:目标检测 -
Data-Driven Feature Tracking for Event Cameras
单位:苏黎世大学-----------关键词:跟踪 -
Integral Neural Networks
单位:华为诺亚(China)
国产4篇,Google3篇,Stanford2篇,艾伦、萨里、苏黎世各1篇;