Object detection Using Detection Transformer (Detr) on custom dataset

Step by step implementation of Object detection Using Detection Transformer (Detr) on custom dataset.
Github: github.com/Aar...
Dataset: universe.robof...
*********************************************************************
For queries: You can comment in comment section or you can mail me at [e-mail]
********************************************************************
DETR stands for "DEtection TRansformer," is a object detection model that uses a transformer architecture.
It was introduced in a research paper titled "End-to-End Object Detection with Transformers," published by researchers from Facebook AI Research (FAIR) in 2020.
CNN extracts features and then send them to transformer for relationship modeling and then obtained output is matches with the ground truth on the picture using bipartite graph matching algorithm.
The features extracted by CNN are flattened and then positional encoding is added to obtain the sequence features which are then fed to transformer encoder.
Each encoder layer contains self attention mechanism and each decoder contains self attention and cross-attention.
#transformers #detr #computervision