YOLOSA: Object detection based on 2D local feature superimposed self-attention

TL;DR

This paper introduces YOLOSA, a real-time object detection model that employs a novel 2D local feature superimposed self-attention mechanism, improving accuracy and efficiency over existing methods.

Contribution

The paper proposes a new self-attention module for feature concatenation, along with an optimized decoupled head and AB-OTA, achieving state-of-the-art results in real-time object detection.

Findings

Achieved up to 49.0% average precision at 71FPS.

Outperformed YOLOv5 by 0.8% to 3.1% in average precision.

Demonstrated improved detection accuracy and inference speed.

Abstract

We analyzed the network structure of real-time object detection models and found that the features in the feature concatenation stage are very rich. Applying an attention module here can effectively improve the detection accuracy of the model. However, the commonly used attention module or self-attention module shows poor performance in detection accuracy and inference efficiency. Therefore, we propose a novel self-attention module, called 2D local feature superimposed self-attention, for the feature concatenation stage of the neck network. This self-attention module reflects global features through local features and local receptive fields. We also propose and optimize an efficient decoupled head and AB-OTA, and achieve SOTA results. Average precisions of 49.0% (71FPS, 14ms), 46.1% (85FPS, 11.7ms), and 39.1% (107FPS, 9.3ms) were obtained for large, medium, and small-scale models built using our proposed improvements. Our models exceeded YOLOv5 by 0.8% -- 3.1% in average precision.