Dynamic Head: Unifying Object Detection Heads with Attentions
This work addresses the challenge of improving object detection performance for computer vision applications by providing a unified approach, though it appears incremental as it builds on existing attention mechanisms.
The paper tackles the problem of unifying object detection heads by proposing a dynamic head framework that combines multiple self-attention mechanisms for scale, spatial, and task awareness, achieving a new state-of-the-art of 54.0 AP on COCO with a ResNeXt-101-DCN backbone and up to 60.6 AP with a transformer backbone and extra data.
The complex nature of combining localization and classification in object detection has resulted in the flourished development of methods. Previous works tried to improve the performance in various object detection heads but failed to present a unified view. In this paper, we present a novel dynamic head framework to unify object detection heads with attentions. By coherently combining multiple self-attention mechanisms between feature levels for scale-awareness, among spatial locations for spatial-awareness, and within output channels for task-awareness, the proposed approach significantly improves the representation ability of object detection heads without any computational overhead. Further experiments demonstrate that the effectiveness and efficiency of the proposed dynamic head on the COCO benchmark. With a standard ResNeXt-101-DCN backbone, we largely improve the performance over popular object detectors and achieve a new state-of-the-art at 54.0 AP. Furthermore, with latest transformer backbone and extra data, we can push current best COCO result to a new record at 60.6 AP. The code will be released at https://github.com/microsoft/DynamicHead.