Research
On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

169,291 papers · 148 categories

Trend · papers per month

75150224299 · Jun 202019922001200920182026
48 results for spectral image

Spectral decoupling improves neural network generalization in medical imaging.

problem Poor generalization of neural networks trained on medical imaging data.
method Spectral decoupling, a regularization technique that encourages learning more features.
result Spectral decoupling increases network robustness and performance on external datasets.

A 3-stage method enhances hyperspectral image classification accuracy.

problem Classifying detailed classes in hyperspectral images with limited labeled data.
method Uses Nested Sliding Window and PCA for spatial consistency, SVM for spectral estimation, and TV model for spatial smoothing.
result Our method outperforms state-of-the-art algorithms, especially in scenarios with small training sets.

SML improves pancreatic mass diagnosis accuracy using CT images.

problem Improving accuracy in pancreatic mass screening using CT imaging.
method Spectral machine learning method trained on 30,000 images, choosing fundamental images based on eigenvectors and removing irrelevant pixels.
result Achieved 94.6% test accuracy in diagnosing 113 patients.

This work proposes a novel autoencoder for fusing visible and infrared images.

problem Challenging task to combine spatial and spectral information from visible and infrared images.
method Spatially constrained adversarial autoencoder with residual architecture and adversarial regularizer.
result Generates a more realistic fused image with enhanced spatial and spectral information.

A new model combines spatial and spectral features for HSI classification.

problem Inefficiency and difficulty in training RNNs for HSI classification.
method Proposes St-SS-pGRU combining shorten RNN, converlusion layer, and parallel-GRU.
result Better performance and robustness in HSI classification.

Proposes a DR method for HSI classification with limited labeled data.

problem High dimensionality, singularity, limited training samples, lack of labeled data, heteroscedasticity, nonlinearity.
method Semi-supervised graph-based, HGF for smoothing, MLP for linear patches, RMLSC and NPMLSC for dissimilarity matrices, optimization for projection matrix.
result Improves HSI classification with limited labeled data.

A new DR method for HSI classification improves accuracy with limited samples.

problem Challenges in DR for HSI classification with limited training samples.
method Graph-based spatial and spectral regularized local scaling cut (SSRLSC).
result Improved classification accuracy compared to spectral-only methods.

DeDUCE efficiently generates realistic counterfactuals for image classifiers.

problem Generating accurate counterfactual explanations for large image classifiers.
method Developed a new algorithm using spectral normalization to efficiently generate counterfactuals.
result Our algorithm consistently produces counterfactuals closer to the original inputs with comparable realism.

This work presents a method to localize targets in hyperspectral images using robust PCA.

problem Localizing targets in hyperspectral images with correlated signatures.
method Modeling HS images as a low-rank plus sparse component, using generalized robust PCA.
result Recovery guarantees and experimental validation show the method's effectiveness.

PixelHop++ improves image classification with a smaller model size.

problem Improving image classification models with smaller sizes.
method Decomposing input tensor, channel-wise Saab transform, successive subspace learning, feature ranking.
result PixelHop++ offers a flexible tradeoff between model size and performance.

Study improves CNN medical image segmentation accuracy and reliability.

problem Over-confident predictions and silent failures in out-of-distribution data.
method Multi-task learning and spectral analysis of CNN feature maps.
result Joint multi-task learning models outperform dedicated models and detect OOD data more accurately.

A new method resolves permutation issues in shuffled linear regression for large-scale applications.

problem Estimating latent features through linear transformation with unknown permutations.
method Spectral matching method to align spectral components of measurement and feature covariances.
result Achieves accurate estimates in shuffled LS and LASSO settings with sufficient samples.

PAC-Bayesian matrix completion with a spectral scaled Student prior offers efficient inference.

problem Matrix completion with underlying low-rank structure.
method Spectral scaled Student prior and PAC-Bayesian bounds.
result Minimax-optimal oracle inequality for model misspecification and general sampling distribution.

New spectral approach improves space-filling designs in high dimensions.

problem Designing high-quality space-filling sample designs in arbitrary dimensions.
method Quantifying space-filling property, connecting spatial and spectral domains, deriving optimal designs, and developing an optimization framework.
result Proposed space-filling spectral designs significantly outperform existing methods, especially in high dimensions.

In spectral clustering and spectral image segmentation, the data is partioned starting from a given matrix of pairwise similarities S. the matrix S is constructed by hand, or learned on a separate training set. In this paper we show how to achieve spectral clustering in unsupervised mode. Our algorithm starts with a se…

2012-07-04abs ↗pdf ↗

Extract coherent structures from Jupiter's clouds and snow events without flow models.

problem Finding coherent structures in fluidic systems from image data.
method Image processing with anisotropic, directed diffusion operator from a directed affinity matrix.
result Demonstrated coherence in Jupiter's clouds and snow events without flow modeling.

Paper proposes a new method to handle spectral variability in hyperspectral unmixing.

problem Spectral variability within endmember classes affects unmixing performance.
method Adaptive bundles and double sparsity to promote sparsity on spectra and classes.
result Successfully determines variable number of classes and estimates their abundances.

Unified analysis for robust PCA with applications to target localization in HS images.

problem Decomposing data matrices into low-rank and sparse components with known dictionaries.
method Unified convex demixing method for entry-wise and column-wise sparse structures, analyzing undercomplete and overcomplete cases.
result Successful recovery of constituent matrices under mild conditions on incoherence, sparsity, and rank.

We introduce a new parameterization method for deep learning layers using spectral tensor train decomposition.

problem Efficiency and stability in deep learning models with weight matrix compression.
method Spectral Tensor Train Parameterization (STTP) of weight matrices.
result Improved compression and training stability in neural networks.

Novel graph-based framework for hyperspectral image classification using superpixels.

problem High classification accuracy with limited labelled data in hyperspectral images.
method Superpixel method for defining local regions, spectral and spatial features extraction, contracted graph representation, semi-supervised classifier.
result Our approach produces accurate classifications with minimal labelled data, outperforming state-of-the-art techniques.

SAGAN improves image generation by leveraging attention across feature locations.

problem Traditional GANs generate details only from local feature maps, limiting image quality.
method SAGAN uses self-attention to model long-range dependencies and spectral normalization for better training.
result SAGAN achieves state-of-the-art results, boosting Inception score and reducing Frechet Inception distance.

A novel framework combines deep metric learning and conditional random field for hyperspectral image classification.

problem Improving classification performance in hyperspectral image processing with limited labeled data.
method Combines spectrum-based deep metric learning and conditional random field. Uses center loss for spectrum-based features and Gaussian edge potentials for pixel-wise classification.
result Demonstrates advantages in classification accuracy and computation cost compared to classical methods.