InverseNet: Benchmarking Operator Mismatch and Calibration Across Compressive Imaging Modalities

TL;DR

This paper introduces InverseNet, a comprehensive benchmark for evaluating the impact of operator mismatch in various compressive imaging systems, revealing significant performance drops in deep learning methods under mismatch conditions.

Contribution

It provides the first cross-modality benchmark for operator mismatch, assessing multiple methods and calibration scenarios across simulated and real data in compressive imaging.

Findings

Deep learning methods lose 10-21 dB under mismatch, negating their advantages.

Performance is inversely related to robustness across modalities.

Operator-conditioned methods recover up to 90% of mismatch losses.

Abstract

State-of-the-art EfficientSCI loses 20.58 dB when its assumed forward operator deviates from physical reality in just eight parameters, yet no existing benchmark quantifies operator mismatch, the default condition in deployed compressive imaging systems. We introduce InverseNet, the first cross-modality benchmark for operator mismatch, spanning CASSI, CACTI, and single-pixel cameras. Evaluating 12 methods under a four-scenario protocol (ideal, mismatched, oracle-corrected, blind calibration) across 27 simulated scenes and 9 real hardware captures, we find: (1) deep learning methods lose 10-21 dB under mismatch, eliminating their advantage over classical baselines; (2) performance and robustness are inversely correlated across modalities (Spearman r_s = -0.71, p < 0.01); (3) mask-oblivious architectures recover 0% of mismatch losses regardless of calibration quality, while operator-conditioned methods recover 41-90%; (4) blind grid-search calibration recovers 85-100% of the oracle bound without ground truth. Real hardware experiments confirm that simulation trends transfer to physical data. Code will be released upon acceptance.