Computer Vision Eccv 2022: 17th European Confe... Apr 2026

: Features like BEVFormer used spatiotemporal transformers to learn unified BEV representations from multi-camera images, which is a critical advancement for autonomous driving perception.

: Frameworks like SimMIM showed that simple random masking strategies could help learn high-quality image representations across various architectures, including ViT and ConvNets. Computer Vision ECCV 2022: 17th European Confe...

Other significant topics explored during the conference included , multimodal learning (combining vision and language), and open-vocabulary object detection . multimodal learning (combining vision and language)

: A simple and effective association method that improved tracking by associating almost every detection box, including those with low scores that were previously discarded. This approach significantly reduced fragmented trajectories and missing objects. Computer Vision ECCV 2022: 17th European Confe...