Quasi-SVD: Learning a Lie-constrained matrix factorisation for real-time imaging

• URL: http://arxiv.org/abs/2607.25967v1
• Date: Tue, 28 Jul 2026 16:47:59 GMT
• Title: Quasi-SVD: Learning a Lie-constrained matrix factorisation for real-time imaging
• Abstract summary: Singular Value Decomposition (SVD) underlies matrix factor tasks across computational imaging.<n>SVD algorithms are inherently sequential, constraining real-time throughput and limit online deployment in clinical pipelines.<n>This study introduces Quasi-SVD, a differentiable, fully parallelized matrix factorization framework for GPUs.
• Score: 6.836244339651614
• License: http://creativecommons.org/licenses/by-sa/4.0/
• Abstract: Singular Value Decomposition (SVD) underlies matrix factorisation tasks across computational imaging, with medical applications increasingly demanding real-time processing. Yet SVD algorithms are inherently sequential, constraining real-time GPU throughput and limit online deployment in clinical pipelines. This study introduces Quasi-SVD, a differentiable, fully parallelized matrix factorization framework for GPUs. Rather than enforcing orthogonality on both factors, it guarantees exact orthogonality for a single Lie-parameterized factor while recovering the remaining components through soft constraints, enabling efficient parallel decomposition without iterative singular-vector optimization. This asymmetric design, provably sufficient for valid factorisation, achieves reconstruction fidelity of SSIM = 0.89-0.94 and accelerates computation by 3-20x relative to cuSOLVER and randomised SVD, enabling throughput above 25 FPS. Performance is evaluated on two medical imaging tasks spanning complementary computational regimes: (1) spatio-temporal background subtraction for ultrasound localisation microscopy, requiring high-dimensional matrix separation, and (2) Mueller matrix polarimetry for neurosurgical tissue characterisation, requiring massive batch processing of small matrices. Across both regimes and multiple imaging instruments, the proposed framework demonstrates robust domain transfer and throughput exceeding 25 FPS at clinical matrix scales, a rate sufficient for live image-guided workflows that classical solvers cannot currently support in these settings. By prioritising downstream reconstruction fidelity over exact spectral recovery, Quasi-SVD makes structured matrix factorisation practical for real-time imaging.

Related papers

• Convolution-Free Holistic Multivariance Decomposition Layer for Efficient Hyperspectral Image Classification Tensor Networks [0.0]
  This work introduces the Holistic Multispectral Decomposition (HMD) framework as a novel, end-to-end differentiable neural network layer.<n>By explicitly separating independent single mode variations from cooperative higher dimensional interactions via learnable, matrix valued supports, proposed HMD-0, HMD-1 and HMD-2 approximants are optimized jointly.<n>These results demonstrate that the HMD framework provides a structurally robust substitute for traditional convolution in multidimensional HS image classification.
  arXiv  Detail & Related papers  (2026-08-17T08:17:46Z)
• Fast two-dimensional tensor-network contraction via subspace iteration [3.5000658943213594]
  Subspace-it CTMRG (SI-CTMRG) is a QR-based projector construction that replaces each large-matrix SVD with an SVD of a much smaller matrix.<n>We demonstrate the efficiency and accuracy of the method for the triangular-lattice Heisenberg antiferromagnetization, reaching state-of-the-art iPEPS results on a single H100 GPU in approximately 10 hours of computation.
  arXiv  Detail & Related papers  (2026-07-16T16:00:54Z)
• VEMamba: Efficient Isotropic Reconstruction of Volume Electron Microscopy with Axial-Lateral Consistent Mamba [10.193877972715667]
  Volume Electron Microscopy (VEM) is crucial for 3D tissue imaging.<n>Existing methods for isotropic reconstruction often suffer from neglecting abundant axial information.<n>We propose VEMamba, an efficient framework for isotropic reconstruction.
  arXiv  Detail & Related papers  (2026-03-01T03:22:42Z)
• Multi-Dimensional Visual Data Recovery: Scale-Aware Tensor Modeling and Accelerated Randomized Computation [51.65236537605077]
  We propose a new type of network compression optimization technique, fully randomized tensor network compression (FCTN)<n>FCTN has significant advantages in correlation characterization and transpositional in algebra, and has notable achievements in multi-dimensional data processing and analysis.<n>We derive efficient algorithms with guarantees to solve the formulated models.
  arXiv  Detail & Related papers  (2026-02-13T14:56:37Z)
• Resolution-Independent Neural Operators for Multi-Rate Sparse-View CT [67.14700058302016]
  Deep learning methods achieve high-fidelity reconstructions but often overfit to a fixed acquisition setup.<n>We propose Computed Tomography neural Operator (CTO), a unified CT reconstruction framework that extends to continuous function space.<n>CTO enables consistent multi-sampling-rate and cross-resolution performance, with on average >4dB PSNR gain over CNNs.
  arXiv  Detail & Related papers  (2025-12-13T08:31:46Z)
• NeuMatC: A General Neural Framework for Fast Parametric Matrix Operation [75.91285900600549]
  We propose textbftextitNeural Matrix Computation Framework (NeuMatC), which elegantly tackles general parametric matrix operation tasks.<n>NeuMatC unsupervisedly learns a low-rank and continuous mapping from parameters to their corresponding matrix operation results.<n> Experimental results on both synthetic and real-world datasets demonstrate the promising performance of NeuMatC.
  arXiv  Detail & Related papers  (2025-11-28T07:21:17Z)
• Non-Intrusive Parametrized-Background Data-Weak Reconstruction of Cardiac Displacement Fields from Sparse MRI-like Observations [0.0]
  We apply the non-intrusive Parametrized Data-Weak (PBDW) approach to 3D cardiac displacement reconstruction from limited MRI-like observations.<n>Our implementation requires only solution snapshots -- no governing equations, assembly routines, or solver access.<n>We demonstrate the effectiveness of the method through validation on a 3D left ventricular model with simulated scar tissue.
  arXiv  Detail & Related papers  (2025-09-18T11:10:24Z)
• Accelerating 3D Photoacoustic Computed Tomography with End-to-End Physics-Aware Neural Operators [74.65171736966131]
  Photoacoustic computed tomography (PACT) combines optical contrast with ultrasonic resolution, achieving deep-tissue imaging beyond the optical diffusion limit.<n>Current implementations require dense transducer arrays and prolonged acquisition times, limiting clinical translation.<n>We introduce Pano, an end-to-end physics-aware model that directly learns the inverse acoustic mapping from sensor measurements to volumetric reconstructions.
  arXiv  Detail & Related papers  (2025-09-11T23:12:55Z)
• Compressive Imaging Reconstruction via Tensor Decomposed Multi-Resolution Grid Encoding [50.54887630778593]
  Compressive imaging (CI) reconstruction aims to recover high-dimensional images from low-dimensional measurements compressed.<n>Existing unsupervised representations may struggle to achieve a desired balance between representation ability and efficiency.<n>We propose Decomposed multi-resolution Grid encoding (GridTD), an unsupervised continuous representation framework for CI reconstruction.
  arXiv  Detail & Related papers  (2025-07-10T12:36:20Z)
• FFT-based Dynamic Subspace Selection for Low-Rank Adaptive Optimization of Large Language Models [49.397861654088636]
  We propose a two-step procedure to approximate SVD/QR-based gradient projections into lower-dimensional spaces.<n>We show that our strategy achieves faster runtime and reduced memory usage by up to $25%$ across different model sizes.
  arXiv  Detail & Related papers  (2025-05-23T14:37:00Z)
• Flattening Singular Values of Factorized Convolution for Medical Images [2.41019965808244]
  Convolutional neural networks (CNNs) have long been the paradigm of choice for robust medical image processing (MIP) Many methods employ factorized convolutional layers to alleviate the burden of limited computational resources. We propose a Singular value equalization generalizer-induced Factorized Convolution (SFConv) to improve the expressive power of factorized convolutions in MIP models.
  arXiv  Detail & Related papers  (2024-03-01T15:30:50Z)
• Orthogonal Matrix Retrieval with Spatial Consensus for 3D Unknown-View Tomography [58.60249163402822]
  Unknown-view tomography (UVT) reconstructs a 3D density map from its 2D projections at unknown, random orientations. The proposed OMR is more robust and performs significantly better than the previous state-of-the-art OMR approach.
  arXiv  Detail & Related papers  (2022-07-06T21:40:59Z)
• Active Phase-Encode Selection for Slice-Specific Fast MR Scanning Using a Transformer-Based Deep Reinforcement Learning Framework [34.540525533018666]
  We propose a light-weight transformer based deep reinforcement learning framework for generating high-quality slice-specific trajectory. The proposed method is roughly 150 times faster and achieves significant improvement in reconstruction accuracy.
  arXiv  Detail & Related papers  (2022-03-11T05:05:09Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.