Semantic features improve satellite sensing over noisy channels
Task-Oriented Semantic Feature Transmission for Multi-Task Satellite Remote Sensing over Low-SNR Channels
Computer Vision and Pattern RecognitionMachine Learning
Summary
Sending satellite images through noisy signals often makes them hard to use for tasks like identifying objects or places. Instead of sending the whole image, the authors suggest sending only important features directly related to these tasks. They created a system that shrinks these features before sending and then fixes them after noisy transmission. Their approach works better than traditional methods, especially when the signal is very weak.
What this means in practice
- •For satellite communication engineers: Enhance satellite data transmission by sending compressed task-specific features instead of full images to improve classification and detection under noisy conditions.
- •For remote sensing software developers: Build lightweight modules that adapt semantic features for better performance in low-quality channel environments during multi-task image analysis.
Authors
Shuoyuan Sun, Hongyu Wang, Mugen Peng, Wenjia Xu
Abstract
Conventional satellite remote sensing transmission follows a reconstruct-then-infer paradigm that optimizes pixel-level fidelity, creating an objective mismatch with downstream tasks such as classification and detection, especially at low SNR. This paper investigates a task-oriented framework that bypasses image reconstruction and directly transmits semantic features extracted by a multitask-pretrained backbone. A lightweight channel adaptation module (CAM) compresses feature dimensionality for bandwidth reduction, and a feature restorer recovers task-relevant structure after channel corruption. With the backbone frozen, the CAM and task-specific downstream heads are jointly optimized with task and feature-level supervision under random-SNR training. Under the adopted AWGN setting, experiments on scene classification and object detection show consistent gains over reconstruction-oriented JSCC baselines across different SNR conditions, with the largest improvements in the low-SNR regime.