Method learns to separate labeled object variations into specified and unspecified factors.
problem Disentangling labeled object variations into specified and unspecified factors.
method Conditional generative model with adversarial training.
result Method can disentangle labeled object variations into specified and unspecified factors.
GMNN combines conditional random fields and graph neural networks for relational data.
problem Semi-supervised object classification in relational data.
method Combines conditional random fields and graph neural networks. Uses variational EM algorithm for training.
result GMNN achieves state-of-the-art results on object classification, link classification, and unsupervised node representation learning.
SoftCVI uses contrastive estimation to infer complex posteriors.
problem Estimating complex posteriors in Bayesian inference.
method Contrastive variational inference with self-generated soft labels.
result SoftCVI outperforms other variational approaches in stability and coverage.
New Variational InfoMax objective improves neural network performance.
problem Optimizing neural networks using Bayesian Inference and Information Bottleneck.
method Derive Variational InfoMax (VIM) objective that maximizes InfoMax directly.
result VIM improves model performance in accuracy, robustness, and representation quality.
Proposes SLCVAE to improve text diversity by self-labeling.
problem KL-Vanishing problem in CVAE for diverse text generation.
method Explicit optimizing objective to guide encoder towards best encoder, using a labeling network.
result Improves text diversity while maintaining comparable accuracy.
This paper investigates semi-supervised hashing methods using variational autoencoders.
problem Semantic hashing with scarce labels.
method Two semi-supervised approaches: joint modeling and pairwise loss.
result The pairwise approach can improve hash quality with many labeled points but degrades with few labels.
We explore the question of whether the representations learned by classifiers can be used to enhance the quality of generative models. Our conjecture is that labels correspond to characteristics of natural data which are most salient to humans: identity in faces, objects in images, and utterances in speech. We propose …
Proposes a model for semi-supervised learning using both labeled and unlabeled data.
problem Semi-supervised learning with limited labeled data.
method Semi-conditional normalizing flow model with conditional coupling layer.
result Model outperforms variational auto-encoders on MNIST dataset.
Efficiently learns from partial labels using variational inference.
problem Learning from noisy and ambiguous partial labels in crowdsourcing.
method Amortized variational inference for probabilistic posterior approximation.
result Achieves state-of-the-art performance in accuracy and efficiency.
New approach models noisy labels without assuming a unique true label.
problem Aggregating noisy labels in ambiguous scenarios.
method Introduces a novel latent distribution assumption.
result The new assumption is more suitable for tasks with ambiguous labels.
CCVAE captures label characteristics in VAEs for better representation learning.
problem Capturing rich label characteristics in VAEs without conflating them with label values.
method Developed CCVAE, a novel VAE model that explicitly captures label characteristics in latent space.
result CCVAE allows for effective and general interventions like smooth traversals and diverse conditional generation.
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.
A new probabilistic model for yes/no crowdsourcing in multi-class classification.
problem Difficulty in discerning among all possible classes in full questions.
method Developed a full probabilistic model for shorter yes/no queries, estimating joint posterior distributions and using Monte Carlo Sampling and Black Box Variational Inference.
result Comparable results to full query crowdsourcing, highlighting the importance of modeling labelers' failures.
In this work we introduce a mixture of GPs to address the data association problem, i.e. to label a group of observations according to the sources that generated them. Unlike several previously proposed GP mixtures, the novel mixture has the distinct characteristic of using no gating function to determine the associati…
This paper refines human labeling as a measurement process, revealing four sources of variation.
problem Systematic variation in human labeling obscures model learning.
method Introduces a statistical framework to decompose labeling outcomes.
result Empirical evidence for four components of labeling variation.
Proposes a method to adapt labels on graphs with few labeled nodes.
problem Domain adaptation for graphs with limited labeled nodes.
method Optimization problem solving label transfer using spectral graph wavelets.
result Method yields satisfactory classification accuracy compared to existing methods.
Paper tackles active learning under human label variation, proposing a new framework.
problem Active learning assumes a single ground truth, ignoring human label variation.
method Survey and propose a conceptual framework for HLV-aware active learning.
result Lay a conceptual foundation for HLV-aware active learning.
BiLA uses variational Bayesian inference to aggregate noisy labels online.
problem Aggregating noisy labels from crowd workers in real-time.
method Variational Bayesian inference and stochastic optimization.
result BiLA reduces label error by at least 10-1.5% points.
OPVI uses operators to optimize variational objectives, improving scalability and approximation quality.
problem Statistical properties of classical variational inference can be undesirable.
method OPVI redefines variational inference using operators to optimize variational objectives.
result OPVI enables data subsampling and variational programs, improving scalability and approximation quality.
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.
Model generates label-dependent paraphrases for NLP tasks.
problem Generating semantically different paraphrases for NLP tasks.
method Deep variational model with label-dependent generation.
result Model improves generative power of paraphrasing models.
New method improves domain generalization by matching object representations.
problem Existing domain generalization methods fail to generalize to unseen domains.
method Proposes matching-based algorithms to match object representations across domains.
result MatchDG algorithm matches ground-truth object representations and improves out-of-domain accuracy.
Paper introduces a new variational objective using Alpha-Beta divergence.
problem Improving variational inference methods for complex distributions.
method Direct optimization of the sAB divergence with two control parameters.
result The sAB divergence framework provides a smooth interpolation and trade-offs between distribution properties.
Variational inference improves training of generative flow networks.
problem Training generative flow networks efficiently and accurately.
method Define variational objectives in terms of KL divergences and optimize convex combinations.
result Variational inference methods can reduce the variance of gradients in training generative flow networks.
Bayesian nonparametric models improve multi-object tracking robustness and flexibility.
problem Challenges in multi-object tracking, including time-dependent cardinality, unordered measurements, and object labeling.
method Robust nonparametric methods for state prior modeling, incorporating dynamic object birth/death and survival.
result Proposed models estimate time-dependent object cardinality, provide object labeling, and identify associated measurements.
AVO improves variational inference by encouraging exploration in latent space.
problem Biasing the true posterior to be unimodal limits the density learned in variational inference.
method Inspired by Annealed Importance Sampling, AVO incorporates energy tempering into the optimization objective.
result AVO facilitates learning by encouraging exploration in latent space, improving robustness and benefits.
Extracts object-centric frames from unlabeled images.
problem Extracting abstract models of 3D objects from visual measurements.
method Viewpoint factorization and dense equivariant labelling neural network.
result Extracts dense object-centric coordinate frames invariant to deformations.
We consider the problem of naming objects in complex, natural scenes containing widely varying object appearance and subtly different names. Informed by cognitive research, we propose an approach based on sharing context based object hypotheses between visual and lexical spaces. To this end, we present the Visual Seman…
M4L-JMF tackles multi-typed objects learning, improving on M3L.
problem Learning from multi-typed objects with diverse features and labels.
method Joint matrix factorization to encode and factorize multi-typed bags and their instances.
result M4L-JMF outperforms existing methods on benchmark datasets.
New f-divergence measures improve robustness in noisy label learning.
problem Improving robustness in learning with noisy labels.
method Derived decoupling property of f-divergence measures under label noise. result Properly defined f-divergence measures are robust with label noise. Proposes a robust VIB approach using soft labels and mutual info estimation.
problem Improving robustness of VIB to adversarial perturbations.
method Refines categorical class information with soft labels from a reference network, relaxes Gaussian posterior assumption.
result Significantly outperforms benchmarked models on MNIST and CIFAR-10.
Bayesian methods improve tracking multiple objects through dynamic dependencies.
problem Tracking multiple objects with time-varying cardinality and unordered measurements.
method Employing Bayesian nonparametric models, specifically dependent Dirichlet and Pitman-Yor processes, for state estimation and Monte Carlo sampling for trajectory learning.
result The proposed methods outperform existing algorithms in estimating the time-varying number of objects and identifying object associations.
Improves least squares classifiers with unlabeled data.
problem Improving supervised classifiers using unlabeled data.
method Block coordinate descent applied to two objective functions.
result Soft-label variant typically outperforms hard-label variant.
The paper introduces multimodal generative models to improve data marginal likelihood.
problem Improving data marginal likelihood in multimodal settings.
method Derives variational bounds on the evidence for multimodal deep generative models, generalizes objectives for different model types, and benchmarks across various datasets.
result Multimodal VAEs excel in image, label, and text datasets with and without weak supervision.
A framework for disentangling class-related and class-independent factors in data.
problem Learning disentangled representations in variational autoencoders.
method Attention mechanism in latent space, mixture models, Bhattacharyya coefficient, semi-supervised training.
result Disentangles class-related and class-independent factors of variation.
Expands weak supervision by allowing partial labels from multiple noisy sources.
problem Creating models without labeled data using heuristic labelers.
method Probabilistic generative model estimating partial label accuracies.
result Improved model accuracy on various tasks (8.6% on text, comparable to zero-shot methods on images).
Paper tackles noisy label distributions in machine learning.
problem Learning from noisy label distributions where true labels are distorted by unknown noise.
method Proposes a probabilistic model with hidden noise parameters and learns it using variational Bayesian methods.
result The proposed model outperforms existing methods in true label estimation.
This paper proposes a new method to select labeled data points using VAEs for active learning.
problem High cost of acquiring labels in supervised machine learning.
method Data-driven approach using Variational Autoencoders (VAEs) to select a diverse core-set in a low-dimensional latent space.
result Improvement in accuracy over related techniques, highlighting the representation power of generative modeling.
Unsupervised learning classifies transient noise in gravitational wave detectors.
problem Transient noise interferes with gravitational wave signals, causing instability.
method Combines variational autoencoder and invariant information clustering.
result Consistent classification with Gravity Spy project labels.
Paper proposes a robust variational inference method for noisy data.
problem Variational inference struggles with random noise in training data.
method Develops a new robust evidence lower bound for variational autoencoders.
result The new objective outperforms the standard evidence lower bound on noisy datasets.
Total variation minimization clusters partially labeled data points.
problem Clustering partially labeled data points in stochastic block models.
method Total variation minimization as a clustering method.
result Total variation minimization allows for accurate clustering under certain model parameters.
Visualizes classification results with class maps.
problem Label bias in classification predictions.
method Class maps reflecting probability, distance, and mislabeling likelihood.
result Insight into classification results and data structure.
Meta-learning neural networks for better clustering representations.
problem Improving clustering performance with appropriate representations.
method Meta-learning method that trains neural networks for representations using VB inference with an infinite Gaussian mixture model.
result The method achieves higher clustering performance than existing methods.
Method separates data into class and style factors using semi-supervised learning.
problem Separating generative factors of data into class and style vectors.
method Independent Vector Variational Autoencoders with semi-supervised learning and independence term.
result Improves classification performance and generation controllability.
The abstract discusses extending learning objectives to measure theory for better generalization.
problem Improving out-of-distribution generalization and weakly-supervised learning.
method Extending variational learning objectives to measures.
result New objectives on measures may lead to practical algorithms.
Expanding on previous work, this note generalizes geometric structures results.
problem Generalizing geometric structures results.
method Generalization to a class of geometric structures including integrable almost-complex structures.
result Main results generalized to a broader class of geometric structures.
DR technique helps deep models learn faster and better.
problem Vanishing gradients and local minima in deep model training.
method DR technique applies penalties on hidden units to improve learning.
result DR improves convergence and generalization in deep neural networks.
There is a rapidly increasing interest in crowdsourcing for data labeling. By crowdsourcing, a large number of labels can be often quickly gathered at low cost. However, the labels provided by the crowdsourcing workers are usually not of high quality. In this paper, we propose a minimax conditional entropy principle to…