Algorithm Improvements based on the Attention Mechanism of yolo-v5

Abstract

In order to further improve the accuracy of mouse species recognition, a yolo-v5 algorithm based on attention mechanism is proposed. The algorithm consists of two parts, one is the yolo-v5 backbone network, and the second part is to add attention mechanisms to the backbone network. It is used to extract global features, and then uses the attention mechanism to give different weights to the extracted features, and finally achieves the purpose of obtaining the features we need. Finally, the softmax classifier is used for classification. Experiments were validated on a homemade mouse dataset. Classification accuracy of 85% and above, respectively. Compared with other algorithms, this algorithm has a good recognition effect and robustness. Rat categories can be identified more accurately.