Improved YOLOv7 for Small and Overlapping Objects Detection

Kaiwen Liang, Weimin Zhang*, Fangxing Li, Zhou Jiang, Di Zhang

*此作品的通讯作者

科研成果: 书/报告/会议事项章节会议稿件同行评审

摘要

Object detection is a crucial task in Deep Learning applications. However, detecting small, overlapping objects in complex scenes has always been challenging. To overcome this challenge, we propose a new algorithm based on YOLOv7, which incorporates a multi-path attention enhancement module in the neck part and a receptive field enhancement module on the feature layer responsible for forecasting small objects before the head. This allows all detection layers to focus more effectively on features that require attention and enhance receptive fields. Our experiments on the PASCAL VOC dataset demonstrate that our proposed approach significantly improves the detection of small and overlapping objects compared to YOLOv7.

源语言英语
主期刊名2023 IEEE 6th International Conference on Pattern Recognition and Artificial Intelligence, PRAI 2023
出版商Institute of Electrical and Electronics Engineers Inc.
153-157
页数5
ISBN(电子版)9798350325485
DOI
出版状态已出版 - 2023
活动6th IEEE International Conference on Pattern Recognition and Artificial Intelligence, PRAI 2023 - Haikou, 中国
期限: 18 8月 202320 8月 2023

出版系列

姓名2023 IEEE 6th International Conference on Pattern Recognition and Artificial Intelligence, PRAI 2023

会议

会议6th IEEE International Conference on Pattern Recognition and Artificial Intelligence, PRAI 2023
国家/地区中国
Haikou
时期18/08/2320/08/23

指纹

探究 'Improved YOLOv7 for Small and Overlapping Objects Detection' 的科研主题。它们共同构成独一无二的指纹。

引用此