GB-LSR: A Fast Local Spectral Image Representation with a Single Global Bandwidth for Continuous Reconstruction and Super-Resolution

2026-06-17 · Source: Takara TLDR - Daily AI Papers · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Computer Vision · Depth: Expert, medium

Summary

GB-LSR (Global-Bandwidth Local Spectral Representation) is a novel fixed-grid local spectral representation designed for continuous image reconstruction and super-resolution. It partitions image domains into non-overlapping square patches, each utilizing coefficients for a truncated Fourier basis derived from shared convolutional-encoder features. A key innovation is a single trainable scalar bandwidth, shared globally across all patches and images, enabling reconstruction at any continuous coordinate with a fixed-size basis contraction independent of image size. In native-reconstruction benchmarks, GB-LSR's main variant surpasses matched-budget LIIF, LTE, and WIRE re-implementations by 2.8-3.6 dB PSNR and 0.11-0.15 LPIPS, while operating at approximately one-quarter of the slowest baseline's inference cost. For arbitrary-scale super-resolution (ASR), it achieves competitive PSNR-Y, running 1.44x faster than LIIF-RDN and 3.25x faster than LTE-SwinIR at x4. Further optimizations yield a 1.77x speedup with 35% lower peak memory.

Key takeaway

For Machine Learning Engineers developing real-time image reconstruction or super-resolution systems, GB-LSR offers a compelling performance-to-cost advantage. You should evaluate this global-bandwidth local spectral representation, as it delivers significantly higher PSNR and LPIPS scores while being substantially faster than current amortized methods. Consider implementing its optimized variants, which provide up to 1.77x speedup and 35% lower peak memory, crucial for resource-constrained deployments.

Key insights

GB-LSR uses a global bandwidth for local spectral image representation, achieving superior performance and speed in continuous reconstruction and super-resolution.

Principles

• A single global bandwidth can suffice for local spectral image representation.
• Fixed-grid local spectral representations enable continuous image reconstruction.
• Basis contraction cost can be independent of image size.

Method

GB-LSR partitions images into non-overlapping square patches, predicting truncated Fourier basis coefficients from shared convolutional-encoder features, using a globally shared trainable scalar bandwidth for continuous reconstruction.

In practice

• Implement a global scalar bandwidth for efficient image representation.
• Consider removing 4-corner local-ensemble averaging for speedup.
• Widen RDN encoder channels (e.g., 64 to 96) for PSNR shift and speed.

Topics

• Image Reconstruction
• Super-Resolution
• Local Spectral Representation
• Fourier Basis
• Global Bandwidth
• PSNR
• LPIPS

Code references

• oucailab/SDANet
• kli8996/HSISR

Best for: Research Scientist, AI Scientist, Machine Learning Engineer, Computer Vision Engineer

Related on AIssential

• GB-LSR: A Fast Local Spectral Image Representation with a Single Global Bandwidth for Continuous Reconstruction and Super-Resolution — A global scalar bandwidth in local spectral representations can achieve superior image reconstruction and super-resolution performance with high efficiency.
• RPC-GS: Gaussian Splatting with native RPC Rendering for Satellite Imagery — RPC-GS improves satellite imagery 3D reconstruction by natively integrating Rational Polynomial Camera models into Gaussian Splatting, avoiding geometric approximations.
• Beyond Visual Fidelity: Benchmarking Super-Resolution Models for Large-Scale Remote Sensing Imagery via Downstream Task Integration — Traditional SR metrics often fail to predict real-world performance in downstream remote sensing tasks.
• GMBFormer: An NDVI-Guided Global Memory Bank Transformer for Urban Green-Space Extraction from Ultra-High-Resolution Imagery — GMBFormer improves urban green-space extraction by decoupling NDVI and using a global memory bank for semantic reuse.
• Adaptive Band Selection for Hyperspectral Classification with Spatially Disjoint Evaluation — Combining supervised spectral ranking with trainable Hard-Concrete gates improves hyperspectral band selection and classification accuracy.
• AnyBand-Diff: A Unified Remote Sensing Image Generation and Band Repair Framework with Spectral Priors — AnyBand-Diff generates physically consistent remote sensing images by integrating spectral priors and multi-scale physical constraints into a diffusion model.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Takara TLDR - Daily AI Papers.