Detects object edges and assigns class labels without pixel-level annotations.
problem Semantic boundary and edge detection with image-level labels.
method Proposes a novel strategy to perform edge detection and class assignment using whole image neural nets and backpropagation.
result High pixel-wise scores indicate semantic boundary locations, suggesting edge labels are not needed during training.
This research shows unsupervised GANs can perform object segmentation without labels.
problem Performing object segmentation without pixel or image-level labels.
method Used large-scale unsupervised GAN models to differentiate foreground from background.
result Demonstrated high-quality saliency masks and new state-of-the-art performance.
A deep model learns to infer fluorescence labels from unlabeled microscopy images.
problem Challenges in obtaining high quality images of cellular structures due to complex environments and label staining limitations.
method Developed a novel deep model using global pixel transformer layers and dense blocks, incorporating multi-scale input strategy.
result Significantly outperforms state-of-the-art methods in fluorescence image prediction tasks.
New method uses image-level and pixel-level annotations for brain tumor segmentation.
problem Challenges in obtaining pixel-level annotations for brain tumor segmentation.
method Proposes a learning-based framework that combines both pixel- and image-level annotations.
result Method's performance in segmentation quality is competitive with traditional fully-supervised approach.
Lazy labels enable deep learning for microscopy segmentation without full annotation.
problem Lack of pixel-wise annotations limits fully supervised learning for bioimage segmentation.
method Lazy labels combined with coarse labels for training a deep neural network.
result Model achieves accurate segmentation with minimal pixel-wise annotations.
Proposes a new contrastive loss for semi-supervised medical image segmentation.
problem Lack of labeled data for medical image segmentation.
method Uses pseudo-labels and a local contrastive loss to learn good local representations.
result Achieved high segmentation performance on public cardiac and prostate datasets.
A deep learning system classifies hyperspectral images using denoising autoencoders and pixel mixtures.
problem Hyperspectral image segmentation and classification challenges.
method Multiple class-based denoising autoencoders, mixed pixel training augmentation, and morphological operations.
result High performance on the Salinas dataset, verified by deep neural network and morphological hole-filling.
Simple CNN achieves page segmentation for historical documents.
problem Page segmentation of handwritten historical document images.
method Proposes a simple CNN architecture trained on raw image pixels for pixel labeling.
result Simple CNN achieves competitive results compared to other deep architectures.
Simplified interactive image segmentation using kNN graphs.
problem Interactive image segmentation with user-provided labels.
method Undirected kNN graphs for label propagation.
result Effective interactive segmentation with significant accuracy.
Augmented CT slices improve liver lesion classification accuracy.
problem Improving pixel-wise classification of hepatic lesions and normal liver tissues.
method Anatomical data augmentation using adjacent CT slices for training a U-net network.
result Improvement of 3% in success rate, 5% in classification accuracy, and 4% in Dice.
Paper proposes a pipeline for weakly supervised learning in object detection and segmentation.
problem Challenges in weakly supervised learning for object detection and segmentation with only image-level labels.
method Curriculum learning pipeline with object localization, filtering, fusing, and pixel labeling.
result State-of-the-art results in multi-label image classification and weakly supervised object detection.
New neural network separates singing voices from music using cross entropy loss.
problem Separating singing voices from music accompaniment.
method Deep Convolutional Neural Network (CNN) trained with Ideal Binary Mask (IBM) and cross entropy loss.
result Proposed CNN outperforms existing systems in MIREX evaluations.
A simple fix helps saliency methods pass sanity checks.
problem Saliency methods fail simple sanity checks.
method Compute saliency maps for all labels, use competition to identify and remove less relevant pixels.
result CGI method is efficient, requires no additional training, and uses only the input and gradient.
Pixel-wise classification, where each pixel is assigned to a predefined class, is one of the most important procedures in hyperspectral image (HSI) analysis. By representing a test pixel as a linear combination of a small subset of labeled pixels, a sparse representation classifier (SRC) gives rather plausible results …
A new method for spotting symbols in CAD images reduces annotation costs and improves accuracy.
problem Challenging task of labeling symbols from CAD drawings.
method Pixel-wise point location via Progressive Gaussian Kernels (PGK) and local offset.
result The proposed method achieves good generalization on real-world CAD images.
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.
Deep learning improves PS pixel selection in SAR interferometry.
problem Selecting persistent scatterer pixels for geophysical parameter estimation in multi-temporal SAR interferometry.
method Proposed two deep learning architectures: CNN-ISS and CLSTM-ISS trained on phase history to classify PS and non-PS pixels.
result CLSTM-ISS outperforms conventional methods in PS pixel selection and classification accuracy.
A new HAR algorithm uses U-Net for pixel-level gesture recognition.
problem Multi-class window problem in traditional HAR methods.
method U-Net network for activity labeling and prediction at each sampling point.
result Highest accuracy and F1-score compared to other methods.
Weakly supervised model segments tumor areas from whole slide images.
problem Training segmentation models on noisy labeled data from whole slide images.
method Online patch sampling and robust KL divergence extension.
result Model successfully segments tumor areas with strong morphological consistency.
A deep neural network improves document binarization accuracy.
problem Binarizing digital documents with historical degradations.
method Combines FCN with primal-dual network for end-to-end training.
result Achieves state-of-the-art binarization on four out of seven datasets.
CAN learns conditional and interventional distributions from unlabeled data.
problem Learning conditional and interventional distributions from unlabeled data.
method CAN framework with LGN and CIGN architectures, equipped with an intervention mechanism.
result CAN generates both interventional and conditional samples without needing the causal graph.
Scene understanding remains a significant challenge in the computer vision community. The visual psychophysics literature has demonstrated the importance of interdependence among parts of the scene. Yet, the majority of methods in computer vision remain local. Pictorial structures have arisen as a fundamental parts-bas…
SaR-SVM-STV improves hyperspectral image classification with shape-adaptive reconstruction and denoising.
problem Classifying hyperspectral images with limited labeled data.
method Shape-adaptive Reconstruction (SaR) for pixel preprocessing, SVM for probability estimation, and Smoothed Total Variation (STV) for denoising.
result SaR-SVM-STV outperforms SVM-STV with fewer labeled data.
Generative model creates realistic scenes from pixel-wise labels.
problem Creating photo-realistic scenes from pixel-wise labels.
method Semantic bottleneck GAN model combining conditional and unconditional generation networks.
result Model outperforms state-of-the-art models in unsupervised image synthesis.
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.
Bayesian method improves segmentation accuracy with noisy labels.
problem Annotation errors in semantic segmentation due to mislabeling and spatial correlations.
method Approximate Bayesian estimation with spatially correlated discrete distributions and variational inference.
result The method achieves performance comparable to clean labels under moderate noise levels.
Improved AI lung ultrasound segmentation using expert confidence values.
problem Label uncertainty in lung ultrasound due to subjective interpretation by radiologists.
method Designing a data annotation protocol capturing expert confidence, training AI on binarized labels with confidence thresholds.
result Improved AI segmentation and better clinical outcomes (e.g., S/F oxygenation ratio estimation, patient readmission prediction).
A novel tracking method for dense honeybee colonies using pixel personality.
problem Tracking large numbers of densely-arranged, interacting objects in a 2D environment.
method Segmentation-based object detection followed by adaptive object recognition through visual appearance.
result Reconstructed ~46% of trajectories in 5 minutes and 71% of tracks for at least 2 minutes.
Deep learning ranks response surfaces for optimal stopping problems in finance.
problem Ranking response surfaces in stochastic control problems.
method Reformulate as image segmentation problem and apply deep learning algorithms.
result Deep learning provides an efficient method for solving optimal stopping problems.
Paper uses RL to fool neural networks with minimal pixel changes.
problem Vulnerability of neural networks to adversarial attacks.
method Deep Q Learning (DQN) to generate adversarial images.
result DQN can fool neural networks with minimal pixel modifications.
Paper tackles domain adaptation without labeled target data.
problem Performing well on an unlabeled target domain using only labeled source data.
method Learning self-supervised tasks on both source and target domains simultaneously.
result Successfully generalizes to the unlabeled target domain.
New unsupervised image translation method detects changes without labeled data.
problem Detecting changes in images without labeled data.
method Affinity-based change priors and weighted loss functions trained on convolutional neural networks.
result Proposed method outperforms state-of-the-art algorithms in detecting changes.
Novel framework for whole-slide blood cell segmentation.
problem Semantic segmentation of whole-slide blood cell images.
method Convolutional encoder-decoder framework with VGG-16 feature extraction.
result Outstanding accuracy (97.18% global accuracy) in classwise segmentation.
Interactive framework identifies insights in power grid maps.
problem Manual identification of insights in power grid maps is laborious and expertise-dependent.
method Proposes an interactive framework using DenseU-Hierarchical VAE to learn and modulate representations.
result Framework outperforms baseline models in identifying and annotating insights.
New neural network learns seismic horizon locations from few images.
problem Automated detection of seismic horizons from small patches is inefficient.
method Multi-resolution U-net with projected loss-function for non-linear regression.
result Network accurately predicts horizon locations even far from known areas.
RTNet uses both deep and pixel-level features for partial domain adaptation.
problem Selecting relevant source samples for knowledge transfer in partial domain adaptation.
method Reinforced Transfer Network (RTNet) with a reinforced data selector (RDS) and domain adaptation model.
result RTNet achieves state-of-the-art performance on benchmark datasets.
The paper proposes a method to learn discriminative multilevel dictionaries for supervised image classification.
problem Improving sparse representation for supervised image classification.
method Learning structured multilevel dictionaries with discriminative constraints for each class, using reconstruction errors of image patches.
result Competitive results compared to state-of-the-art methods on texture image classification.
Label assignment problems with large state spaces are important tasks especially in computer vision. Often the pairwise interaction (or smoothness prior) between labels assigned at adjacent nodes (or pixels) can be described as a function of the label difference. Exact inference in such labeling tasks is still difficul…
Proposes a strategy to train models with minimal labeled data.
problem Scarce and expensive labeled data for medical tasks.
method Recursive training strategy to use image-level annotations for pixel-level segmentation.
result Improved segmentation of intracranial hemorrhage in CT scans.
RankSEG-RMA improves semantic segmentation efficiency and applicability.
problem Inconsistent or suboptimal semantic segmentation results due to argmax or thresholding.
method Developed RankSEG-RMA using reciprocal moment approximation to optimize Dice and IoU metrics.
result RankSEG-RMA reduces computational complexity to O(d) while maintaining comparable performance.
Study fusion methods for financial image views to improve robustness against attacks.
problem Improving robustness of financial image views for next-day direction prediction.
method Same-source multi-view learning with early fusion and late fusion, using OHLCV and technical-indicator views, and evaluating pixel-space L-infinity attacks.
result Early fusion can suffer negative transfer under noisy settings, while late fusion is more reliable once labels stabilize.
Proposes a method to train deep neural networks robust to label noise in remote sensing images.
problem Training deep neural networks on datasets with inaccurate labels leads to overfitting and poor performance.
method Uses entropic optimal transport to learn robust deep neural networks.
result Empirically demonstrates superior performance compared to state-of-the-art methods on remote sensing datasets.
The study examines instance label stability in MIL classifiers trained on global image annotations.
problem Instance labels in MIL classifiers can be unstable, leading to incorrect fine-grained annotations.
method Investigated instance stability on 5 datasets, proposing an unsupervised measure.
result A performance-stability trade-off can be made when comparing MIL classifiers.
Weakly supervised learning for histopathology disease localization.
problem Locating abnormal cells or single cells in histopathology slides.
method Pre-trained deep convolutional networks, feature embedding, top instances and negative evidence.
result Comparable performance to strong annotations on lymph node metastases detection challenge.
Neural network classifies breast cancer lesions using global and local image features.
problem Classifying breast cancer lesions in medical images with high resolution and small regions of interest.
method Proposes a neural network that combines global saliency maps and local patches for pixel-level saliency maps.
result Achieves radiologist-level performance in screening mammography interpretation.
A new method learns object representations from motion in slot representations.
problem Unsupervised object extraction from low-level visual data.
method Contrastive learning in slot representations, focusing on moving objects and distinct entities.
result Introduced a new evaluation metric to measure diversity of slot vectors.
Paper proposes a method to generate instance labels from weakly supervised data.
problem Weakly supervised instance labeling in medical image analysis.
method Uses multiple instance learning (MIL) and knowledge distillation to generate instance-level predictions.
result Significantly outperforms state-of-the-art MIL methods in instance-level prediction.
Proposes a new model for unsupervised clustering with latent variables.
problem The challenge of unsupervised clustering in machine learning.
method Clustered Generator Model with continuous and discrete latent variables.
result Achieves competitive unsupervised clustering accuracy and disentangled latent representations.