mlsanityData to 7 Oct 2026

Winner-takes-all learners are geometry-aware conditional density estimators

arXiv 2406.04706 ↗First posted 7 Jun 2024cs.LG cs.NE eess.SPcited by 6

Not on the map. Why?

Abstract (arXiv)

Winner-takes-all training is a simple learning paradigm, which handles ambiguous tasks by predicting a set of plausible hypotheses. Recently, a connection was established between Winner-takes-all training and centroidal Voronoi tessellations, showing that, once trained, hypotheses should quantize optimally the shape of the conditional distribution to predict. However, the best use of these hypotheses for uncertainty quantification is still an open question. In this work, we show how to leverage the appealing geometric properties of the Winner-takes-all learners for conditional density estimation, without modifying its original training scheme. We theoretically establish the advantages of our novel estimator both in terms of quantization and density estimation, and we demonstrate its competitiveness on synthetic and real-world datasets, including audio data.

Key facts. Cited by 6 papers in the map (data through 7 Oct 2026). Builds on 26 of its 63 references that are in the map. First posted on arXiv 7 Jun 2024.

Builds on

26 of 63 references.

References · 26

  1. 2017Attention Is All You Needcited by 59,932
  2. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  3. 2015Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learningcited by 4,416
  4. 2016Simple and Scalable Predictive Uncertainty Estimation using Deep Ensemblescited by 3,888
  5. 2013Generating Sequences With Recurrent Neural Networkscited by 549
  6. 2018Multimodal Trajectory Predictions for Autonomous Driving using Deep Convolutional Networkscited by 289
  7. 2022On the Pitfalls of Heteroscedastic Uncertainty Estimation with Probabilistic Neural Networkscited by 119
  8. 2016Learning in an Uncertain World: Representing Ambiguity Through Multiple Hypothesescited by 96
  9. 2019Overcoming Limitations of Mixture Density Networks: A Sampling and Fitting Framework for Multimodal Future Predictioncited by 96
  10. 2019Generating Multiple Hypotheses for 3D Human Pose Estimation with Mixture Density Networkcited by 92
  11. 2018Sound Event Localization and Detection of Overlapping Sources Using Convolutional Recurrent Neural Networkscited by 84
  12. 2016Stochastic Multiple Choice Learning for Training Diverse Deep Ensemblescited by 77
14 more
  1. 2018Improving Regression Performance with Distributional Lossescited by 66
  2. 2021A Survey of Sound Source Localization with Deep Learning Methodscited by 45
  3. 2021Divide-and-Conquer for Lane-Aware Diverse Trajectory Predictioncited by 37
  4. 2019DMCL: Distillation Multiple Choice Learning for Multimodal Action Recognitioncited by 30
  5. 2017Direction of arrival estimation for multiple sound sources using convolutional recurrent neural networkcited by 27
  6. 2017Confident Multiple Choice Learningcited by 24
  7. 2018Structural Consistency and Controllability for Diverse Colorizationcited by 15
  8. 2023Resilient Multiple Choice Learning: A learned scoring scheme with application to audio scene analysiscited by 12
  9. 2019Localization, Detection and Tracking of Multiple Moving Sound Sources with a Convolutional Recurrent Neural Networkcited by 9
  10. 2016Theoretical Analysis of the k-Means Algorithm - A Surveycited by 7
  11. 2021PILOT: Introducing Transformers for Probabilistic Sound Event Localizationcited by 7
  12. 2022Voronoi Density Estimator for High-Dimensional Data: Computation, Compactification and Convergencecited by 7
  13. 2021Exploiting Attention-based Sequence-to-Sequence Architectures for Sound Event Localizationcited by 6
  14. 2022An Efficient and Continuous Voronoi Density Estimatorcited by 3

Built on by

Cited by year

Citing papers per year: 2024: 1; 2025: 5.

6 papers.

  1. 2024Annealed Multiple Choice Learning: Overcoming limitations of Winner-takes-all with annealingcited by 12
  2. 2025From Samples to Scenarios: A New Paradigm for Probabilistic Forecastingcited by 8
  3. 2025Winner-takes-all for Multivariate Probabilistic Time Series Forecastingcited by 4
  4. 2025TimePre: Bridging Accuracy, Efficiency, and Stability in Probabilistic Time-Series Forecastingcited by 3
  5. 2025Multiple Choice Learning of Low-Rank Adapters for Language Modelingcited by 2
  6. 2025Conditional Distribution Quantization in Machine Learningcited by 0