SUPER Decoder Block for Reconstruction-Aware U-Net Variants

TL;DR

The paper introduces SUPER, a novel decoder block for U-Net variants that leverages wavelet perfect reconstruction to improve high-frequency detail recovery in inverse problems, enhancing segmentation and denoising performance.

Contribution

It proposes SUPER, a flexible, reconstruction-aware decoder block that prevents information loss and boosts high-frequency detail recovery in U-Net architectures without rigid constraints.

Findings

Significantly improves thin-crack segmentation, especially for cracks narrower than 4 px.

Achieves moderate PSNR gains in smartphone image denoising, demonstrating robustness.

Enhances representational diversity and high-frequency fidelity across diverse tasks.

Abstract

Skip-connected encoder-decoder architectures (U-Net variants) are widely adopted for inverse problems but still suffer from information loss, limiting recovery of fine high-frequency details. We present Selectively Suppressed Perfect Reconstruction (SUPER), which exploits the perfect reconstruction (PR) property of wavelets to prevent information degradation while selectively suppressing (SS) redundant features. Free from rigid framelet constraints, SUPER serves as a plug-and-play decoder block for diverse U-Net variants, eliminating their intrinsic reconstruction bottlenecks and enhancing representational richness. Experiments across diverse crack benchmarks, including state-of-the-art (SOTA) models, demonstrate the structural potential of the proposed SUPER Decoder Block. Maintaining comparable computational cost, SUPER enriches representational diversity through increased parameterization. In small-scale in-domain experiments on the CrackVision12K dataset, SUPER markedly improves thin-crack segmentation performance, particularly for cracks narrower than 4 px, underscoring its advantage in high-frequency dominant settings. In smartphone image denoising on SIDD, where low-frequency components prevail, SUPER still achieves a moderate gain in PSNR, confirming its robustness across low- and high-frequency regimes. These results validate its plug-and-play generality across U-Net variants, achieving high-frequency fidelity and global coherence within a unified, reconstruction-aware framework.