An efficient target re-recognition method combining EfficientNet and MeMOT
2023 3rd International Conference on Electronic Information Engineering and Computer (EIECT)(2023)
Abstract
With the wide application of deep learning techniques in computer vision, target re-recognition has become a popular research direction. Traditional target re-recognition methods face many challenges when dealing with target matching in complex scenes and multi-camera fields of view, such as view angle changes, lighting condition differences and occlusion. In order to better cope with these challenges, this paper proposes a novel target re-recognition method combining EfficientNet and MeMOT, referred to as EN-ME. EfficientNet, as a highly efficient deep neural network, is able to extract rich features from images; while MeMOT provides strong spatio-temporal memory support for target tracking. By combining the two, our approach not only accurately matches targets, but also enables efficient target tracking in real-time video streams. We conduct experiments on several publicly available datasets such as Market-1501, DukeMTMC-reID and CUHK03. The experimental results show that the EN-ME method performs better in terms of mAP and Rank-l accuracy. In addition, we analyze the robustness, computational efficiency and scalability of the model.
MoreTranslated text
Key words
Target re-identification,EfficientNet,MeMOT,Deep learning
AI Read Science
Must-Reading Tree
Example
![](https://originalfileserver.aminer.cn/sys/aminer/pubs/mrt_preview.jpeg)
Generate MRT to find the research sequence of this paper
Chat Paper
Summary is being generated by the instructions you defined