Facebook AI Research applies Transformer architecture to streamline object detection models

http://feedproxy.google.com/~r/venturebeat/SZYF/~3/b01YWzN90v8/

A team of 6 members of Facebook AI Research (FAIR) tapped the popular Transformer neural network architecture to create end-to-end object detection AI, an approach they claim streamlines the creation of object detection models and reduces the need for handcrafted components. Named Detection Transformer, aka DETR, the model can recognize objects in an image in a single pass all at once.  » ….