mlsanityData to 7 Oct 2026

SUPER Module for Detail-Sensitive and Cost-Efficient U-Net Variant Decoders

arXiv 2511.11015 ↗First posted 14 Nov 2025cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Skip-connected U-Net variants are widely used for dense inverse problems, yet their decoders commonly recover resolution through spatial upscaling, which can blur or distort fine structures. Wavelet transforms provide an explicit perfect-reconstruction path, but prior wavelet networks often couple this property to wavelet-specific architectures or fixed output representations. We propose Selectively Suppressed Perfect Reconstruction (SUPER), a wavelet-domain decoder formulation that replaces unstructured spatial refinement with bounded frequency suppression. In its idealized equal-channel form, SUPER reduces to exact reconstruction when the suppression term is zero; in practical U-Net decoders, it provides a projected-subspace reconstruction fallback with learnable task-specific suppression. The resulting SUPER Module is a structurally plug-in decoder block: it replaces the upsampling/refinement stage of an existing U-Net-like decoder, while leaving the encoder and task head unchanged. We evaluate SUPER on monocular depth estimation, thin-crack segmentation, and smartphone image denoising. On iDisc depth estimation, SUPER improves the host model's edge AUC from 26.59% to 34.45% while reducing decoder MACs from 12.69G to 4.36G. On FACS-Net crack segmentation, SUPER improves average IoU and the extreme 0-2px crack regime. On SIDD denoising, where high-frequency enhancement is less directly rewarded, SUPER preserves PSNR/SSIM while reducing decoder MACs by 66.6%. These results support SUPER as a practical wavelet-domain suppression module for improving the observed detail-cost trade-off of U-Net-like decoders.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 26 of its 48 references that are in the map. First posted on arXiv 14 Nov 2025.

Builds on

26 of 48 references.

References · 26

  1. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
  2. 2021Restormer: Efficient Transformer for High-Resolution Image Restorationcited by 1,618
  3. 2018On the Spectral Bias of Neural Networkscited by 1,321
  4. 2022Simple Baselines for Image Restorationcited by 701
  5. 2020AdaBins: Depth Estimation using Adaptive Binscited by 596
  6. 20193D Packing for Self-Supervised Monocular Depth Estimationcited by 437
  7. 2020clDice -- A Novel Topology-Preserving Loss Function for Tubular Structure Segmentationcited by 286
  8. 2020Focal Frequency Loss for Image Reconstruction and Synthesiscited by 204
  9. 2018Multi-level Wavelet-CNN for Image Restorationcited by 184
  10. 2019Towards Understanding the Spectral Bias of Deep Learningcited by 173
  11. 2019CARAFE: Content-Aware ReAssembly of FEaturescited by 119
  12. 2021Boosting Monocular Depth Estimation Models to High-Resolution via Content-Adaptive Multi-Resolution Mergingcited by 104
14 more
  1. 2022Attention Attention Everywhere: Monocular Depth Prediction with Skip Attentioncited by 92
  2. 2023iDisc: Internal Discretization for Monocular Depth Estimationcited by 71
  3. 2019Wavelet Domain Style Transfer for an Effective Perception-distortion Tradeoff in Single Image Super-Resolutioncited by 30
  4. 2022FADE: Fusing the Assets of Decoder and Encoder for Task-Agnostic Upsamplingcited by 24
  5. 2023Catch Missing Details: Image Reconstruction with Frequency Augmented Variational Autoencodercited by 18
  6. 2024CascadedGaze: Efficiency in Global Context Extraction for Image Restorationcited by 16
  7. 2024Addressing Spectral Bias of Deep Neural Networks by Multi-Grade Deep Learningcited by 14
  8. 2023Towards Building More Robust Models with Frequency Biascited by 12
  9. 2025A Comprehensive Review of U-Net and Its Variants: Advances and Applications in Medical Image Segmentationcited by 10
  10. 2021How Convolutional Neural Networks Deal with Aliasingcited by 9
  11. 2022Multi-resolution Monocular Depth Map Fusion by Self-supervised Gradient-based Compositioncited by 8
  12. 2024Hybrid-Segmentor: A Hybrid Approach to Automated Fine-Grained Crack Segmentation in Civil Infrastructurecited by 7
  13. 2025Augmenting Perceptual Super-Resolution via Image Quality Predictorscited by 7
  14. 2025A Survey of Deep Learning Video Super-Resolutioncited by 5

Built on by

Not cited yet.