Paper
23 January 2024 Deep compression for real-time high-precision SAR image ship detection
Yue Yang, Zhuqing Yuan, Shuangcai Liu, Wenyu Sun, Yongpan Liu, Sheng Zhang
Author Affiliations +
Proceedings Volume 12978, Fourth International Conference on Geology, Mapping, and Remote Sensing (ICGMRS 2023); 129781K (2024) https://doi.org/10.1117/12.3021068
Event: 2023 4th International Conference on Geology, Mapping and Remote Sensing (ICGMRS 2023), 2023, wuhan, China
Abstract
Synthetic aperture radar(SAR) ship detection plays an important role in ship dispatching, battlefield dynamic tracking and other applications, which all require real-time inference. High spatial resolution of SAR images means a large amount of data and a large computation cost, which makes it difficult to realize real-time inference in hardware with limited resources. Therefore, we propose a new model compression scheme to learn a slim ship detector named Lite-YOLOv4. By modifying network structure to be lightweight and improving polarization-based channel pruning, we generate a compact model. In order to further compress the model, a progressive training-based mixed-precision quantization is proposed to simplify the model bit-representation, reduce its storage requirements and reduce the amount of computation. Finally, the model performance is restored by a self-designed target-guided module with attention mechanism. A lot of experiments were carried out on SSDD, which verify that our method is advanced compared with other CNN-based algorithms. The proposed detector with 0.64MB parameters, achieves 96.6% AP with the calculation cost of 36.11 bitGOPs, in which bitGOPs is 70.9% lower than SOTA, parameter storage is 46.7% lower and AP is 2% higher accuracy.
(2024) Published by SPIE. Downloading of the abstract is permitted for personal use only.
Yue Yang, Zhuqing Yuan, Shuangcai Liu, Wenyu Sun, Yongpan Liu, and Sheng Zhang "Deep compression for real-time high-precision SAR image ship detection", Proc. SPIE 12978, Fourth International Conference on Geology, Mapping, and Remote Sensing (ICGMRS 2023), 129781K (23 January 2024); https://doi.org/10.1117/12.3021068
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
Back to Top