Quantum-Inspired Robust and Scalable SAR Object Classification

TL;DR

This paper investigates tensor networks for SAR image classification, demonstrating their robustness to noise and data poisoning while maintaining a compact model size suitable for edge deployment.

Contribution

It introduces the application of tensor networks to SAR object classification, highlighting their robustness and efficiency advantages over traditional neural networks.

Findings

Tensor networks show resilience to data poisoning in SAR classification.

Tensor networks enable smaller, more efficient models without sacrificing accuracy.

The study provides insights into deploying robust models on edge devices.

Abstract

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.