SatReg: Regression-based Neural Architecture Search for Lightweight Satellite Image Segmentation

TL;DR

SatReg is a regression-based framework that efficiently searches for lightweight satellite image segmentation models optimized for edge devices, balancing accuracy, latency, and power.

Contribution

Introduces a hardware-aware, surrogate-model-based neural architecture search method tailored for remote-sensing segmentation on edge platforms.

Findings

Surrogate models accurately predict mIoU, latency, and power for different architectures.

Selected architecture variables significantly influence accuracy and hardware costs.

The approach enables rapid, near-optimal architecture selection without exhaustive search.

Abstract

As Earth-observation workloads move toward onboard and edge processing, remote-sensing segmentation models must operate under tight latency and energy constraints. We present SatReg, a regression-based hardware-aware tuning framework for lightweight remote-sensing segmentation on edge platforms. Using CM-UNet as the teacher architecture, we reduce the search space to two dominant width-related variables, profile a small set of student models on an NVIDIA Jetson Orin Nano, and fit low-order surrogate models for mIoU, latency, and power. Knowledge distillation is used to efficiently train the sampled students. The learned surrogates enable fast selection of near-optimal architecture settings for deployment targets without exhaustive search. Results show that the selected variables affect task accuracy and hardware cost differently, making reduced-space regression a practical strategy for adapting hybrid CNN-Mamba segmentation models to future space-edge systems.