DETR ResNet-101

JSON →
facebook vision
image

End-to-end object detection model using a transformer with a deeper ResNet-101 backbone for higher accuracy.

vision
releasedMay 2020