D^2ETR: Decoder-Only DETR with Computationally Efficient Cross-Scale Attention
This work addresses computational efficiency for object detection in computer vision, representing an incremental improvement over existing DETR-based methods.
The paper tackles the high computational cost and slow convergence of DETR by proposing D^2ETR, a decoder-only detector with a novel cross-scale attention module, which achieves lower complexity and higher accuracy on the COCO benchmark compared to DETR and its variants.
DETR is the first fully end-to-end detector that predicts a final set of predictions without post-processing. However, it suffers from problems such as low performance and slow convergence. A series of works aim to tackle these issues in different ways, but the computational cost is yet expensive due to the sophisticated encoder-decoder architecture. To alleviate this issue, we propose a decoder-only detector called D^2ETR. In the absence of encoder, the decoder directly attends to the fine-fused feature maps generated by the Transformer backbone with a novel computationally efficient cross-scale attention module. D^2ETR demonstrates low computational complexity and high detection accuracy in evaluations on the COCO benchmark, outperforming DETR and its variants.