Energy-based Out-of-distribution Detection
Explore this paper's citation graph
Summary
This work proposes a unified framework for OOD detection that uses an energy score, and shows that energy scores better distinguish in- and out-of-distribution samples than the traditional approach using the softmax scores.
- Type
- preprint
- Published
- 2020-10-08
- Cited by
- 2,139
- References
- 52
- Access
- Open access
- OpenAlex
- https://openalex.org/W3092527263
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:222208700
Keywords
Softmax function, Overconfidence effect, Confidence interval, Artificial intelligence, Computer science
References
- Efficient Learning of Deep Boltzmann Machines
- LSUN: Construction of a Large-scale Image Dataset using Deep Learning with Humans in the Loop
- TurkerGaze: Crowdsourcing Saliency with Webcam based Eye Tracking
- Stochastic Backpropagation and Approximate Inference in Deep Generative Models
- Deep neural networks are easily fooled: High confidence predictions for unrecognizable images
- A Learning Algorithm for Boltzmann Machines
- Describing Textures in the Wild
- A Family of Nonparametric Density Estimation Algorithms
- 80 Million Tiny Images: A Large Data Set for Nonparametric Object and Scene Recognition
- A Tutorial on Energy-Based Learning
- A Theory of Generative ConvNet
- Reading Digits in Natural Images with Unsupervised Feature Learning
- Synthesizing Dynamic Patterns by Spatial-Temporal Generative ConvNet
- SGDR: Stochastic Gradient Descent with Warm Restarts
- Cooperative Training of Descriptor and Generator Networks
- A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks
- A Unified Energy-Based Framework for Unsupervised Learning
- On Calibration of Modern Neural Networks
- Places: A 10 Million Image Database for Scene Recognition
- Confidence estimation in Deep Neural networks via density modelling
Cited by
- Robust Out-of-distribution Detection for Neural Networks
- Hybrid Discriminative-Generative Training via Contrastive Learning
- Informative Outlier Matters: Robustifying Out-of-distribution Detection Using Outlier Mining
- Variational (Gradient) Estimate of the Score Function in Energy-based Latent Variable Models
- Entropic Out-of-Distribution Detection
- Exploring Vicinal Risk Minimization for Lightweight Out-of-Distribution Detection
- Perfect Density Models Cannot Guarantee Anomaly Detection
- Analyzing and Improving Generative Adversarial Training for Generative Modeling and Out-of-Distribution Detection
- Energy-based Out-of-distribution Detection for Multi-label Classification
- Unsupervised Energy-based Out-of-distribution Detection using Stiefel-Restricted Kernel Machine
- Task-Agnostic Out-of-Distribution Detection Using Kernel Density Estimation
- Iterative human and automated identification of wildlife images
- MOS: Towards Scaling Out-of-distribution Detection for Large Semantic Space
- Elsa: Energy-based Learning for Semi-supervised Anomaly Detection
- MOOD: Multi-level Out-of-distribution Detection
- Contrastive Out-of-Distribution Detection for Pretrained Transformers
- AutoEval: Are Labels Always Necessary for Classifier Accuracy Evaluation?
- Entropic Out-of-Distribution Detection: Seamless Detection of Unknown Examples
- Towards Consistent Predictive Confidence through Fitted Ensembles
- Energy-based Unknown Intent Detection with Data Manipulation
Related papers
- Exploring Alternatives to Softmax Function
- Two Experiments in Reducing Overconfidence in Spreadsheet Development
- Investor Overconfidence in Experimental Asset Markets across Market States
- Effectiveness of MPC-friendly Softmax Replacement
- Improved softmax loss for deep learning-based face and expression recognition
- A softmax classifier for high-precision classification of ultrasonic similar signals.