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