Search papers, labs, and topics across Lattice.
This paper investigates the use of tensor networks for SAR image classification, focusing on their robustness to noise and data poisoning, as well as their potential for model size reduction. The authors evaluate tensor networks' performance in object classification tasks, contrasting them with conventional neural networks. Results show that tensor networks effectively balance robustness and model efficiency in SAR image classification.
Tensor networks offer a surprisingly robust and efficient alternative to traditional neural networks for classifying noisy SAR imagery, even under data poisoning attacks.
SAR image classification naturally has to deal with huge noise and a high dynamic range particularly requiring robust classification models. Additionally, the deployment of these models on edge devices, such as drones and military aircraft, requires a careful balance between model size and classification accuracy. This study explores the potential of tensor networks to meet these robustness requirements, specifically evaluating their resilience to data poisoning. Unlike previous works that concentrated on conventional neural networks for SAR object detection, this research focuses on the robustness and model reduction capabilities of tensor networks in object classification. Our findings indicate that tensor networks are adept at addressing both the challenges of robustness and the need for model efficiency, thereby contributing valuable insights to the ongoing discourse in radar applications and deep learning methodologies in general.
