Discriminant Distance-Aware Representation on Deterministic Uncertainty Quantification Methods
Not on the map. Why?
Abstract (arXiv)
Uncertainty estimation is a crucial aspect of deploying dependable deep learning models in safety-critical systems. In this study, we introduce a novel and efficient method for deterministic uncertainty estimation called Discriminant Distance-Awareness Representation (DDAR). Our approach involves constructing a DNN model that incorporates a set of prototypes in its latent representations, enabling us to analyze valuable feature information from the input data. By leveraging a distinction maximization layer over optimal trainable prototypes, DDAR can learn a discriminant distance-awareness representation. We demonstrate that DDAR overcomes feature collapse by relaxing the Lipschitz constraint that hinders the practicality of deterministic uncertainty methods (DUMs) architectures. Our experiments show that DDAR is a flexible and architecture-agnostic method that can be easily integrated as a pluggable layer with distance-sensitive metrics, outperforming state-of-the-art uncertainty estimation methods on multiple benchmark problems.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 32 of its 48 references that are in the map. First posted on arXiv 20 Feb 2024.
Builds on
32 of 48 references.
References · 32
- 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 11,979
- 2017On Calibration of Modern Neural Networkscited by 4,860
- 2015Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learningcited by 4,416
- 2018Spectral Normalization for Generative Adversarial Networkscited by 1,618
- 2019Can You Trust Your Model's Uncertainty? Evaluating Predictive Uncertainty Under Dataset Shiftcited by 1,274
- 2021A Survey of Uncertainty in Deep Neural Networkscited by 675
- 2020Simple and Principled Uncertainty Estimation with Deterministic Deep Learning via Distance Awarenesscited by 389
- 2019An Evaluation Dataset for Intent Classification and Out-of-Scope Predictioncited by 378
- 2018Analyzing Inverse Problems with Invertible Neural Networkscited by 239
- 2021Adaptive Prototype Learning and Allocation for Few-Shot Segmentationcited by 207
- 2021Deep Deterministic Uncertainty: A Simple Baselinecited by 200
- 2020Contrastive Training for Improved Out-of-Distribution Detectioncited by 186
20 more
- 2020Posterior Network: Uncertainty Estimation without OOD Samples via Density-Based Pseudo-Countscited by 163
- 2019Practical Deep Learning with Bayesian Principlescited by 158
- 2019A General Framework for Uncertainty Estimation in Deep Learningcited by 119
- 2022Plex: Towards Reliability using Pretrained Large Model Extensionscited by 102
- 2021On Feature Collapse and Deep Kernel Learning for Single Forward Pass Uncertaintycited by 87
- 2019Sampling-free Epistemic Uncertainty Estimation Using Approximated Variance Propagationcited by 74
- 2019Out-of-domain Detection for Natural Language Understanding in Dialog Systemscited by 62
- 2021On the Practicality of Deterministic Epistemic Uncertaintycited by 62
- 2019Hybrid Models with Deep and Invertible Featurescited by 41
- 2018Uncertainty in the Variational Information Bottleneckcited by 39
- 2017Distance-based Confidence Score for Neural Network Classifierscited by 30
- 2019Entropic Out-of-Distribution Detectioncited by 21
- 2020Modern Monte Carlo Methods for Efficient Uncertainty Quantification and Propagation: A Surveycited by 21
- 2022Latent Discriminant deterministic Uncertaintycited by 17
- 2020A Simple Framework for Uncertainty in Contrastive Learningcited by 16
- 2021Contrastive Prototype Learning with Augmented Embeddings for Few-Shot Learningcited by 15
- 2020The Hidden Uncertainty in a Neural Networks Activationscited by 14
- 2019Towards Open Intent Discovery for Conversational Textcited by 13
- 2021Natural Posterior Network: Deep Bayesian Uncertainty for Exponential Family Distributionscited by 13
- 2023On the Quantification of Image Reconstruction Uncertainty without Training Datacited by 1
Built on by
Citing papers per year: 2025: 2.
2 papers.