Detection Transformer
This is an implementation of Detection transformer that utilizes the attention mechanism of a transformer as used in natural language processing and natural language understanding task which allows the model to focus on specific region of the model and then not only plot the bounding box around it but also to efficiently detect the object in that area.
This is a one-shot detector with minimal to almost no post-processing steps which makes it highly efficient and its attention mechanism provides it the ability to detect even the smallest object in the image. This capabilities can be of due use for in scenarios like cancel cell detection, or where the object under study has a significant smaller size.
