BAR: Bayesian Activity Recognition using variational inference

2018/11/08 by Ranganath Krishnan, Krishnan, Ranganath, Mahesh Subedar +3 · 1 citation
Computer Science · Mathematics · #Adversarial Robustness in Machine Learning #Anomaly Detection Techniques and Applications #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Human Pose and Action Recognition #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Neural and Evolutionary Computing (cs.NE) #cs.CV #cs.LG #cs.NE #stat.ML

paper · pdf · doi:10.48550/arxiv.1811.03305

openalex publication_date 2018/11/08 · arxiv created 2018/12/01 · arxiv updated 2018/12/04 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Uncertainty estimation in deep neural networks is essential for designing reliable and robust AI systems. Applications such as video surveillance for identifying suspicious activities are designed with deep neural networks (DNNs), but DNNs do not provide uncertainty estimates. Capturing reliable uncertainty estimates in safety and security critical applications will help to establish trust in the AI system. Our contribution is to apply Bayesian deep learning framework to visual activity recognition application and quantify model uncertainty along with principled confidence. We utilize the stochastic variational inference technique while training the Bayesian DNNs to infer the approximate posterior distribution around model parameters and perform Monte Carlo sampling on the posterior of model parameters to obtain the predictive distribution. We show that the Bayesian inference applied to DNNs provide reliable confidence measures for visual activity recognition task as compared to conventional DNNs. We also show that our method improves the visual activity recognition precision-recall AUC by 6.2% compared to non-Bayesian baseline. We evaluate our models on Moments-In-Time (MiT) activity recognition dataset by selecting a subset of in- and out-of-distribution video samples.

Citations

• UCF101: A Dataset of 101 Human Actions Classes From Videos in The Wild
• Black Box Variational Inference
• YouTube-8M: A Large-Scale Video Classification Benchmark
• Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics
• The Kinetics Human Action Video Dataset
• Flipout: Efficient Pseudo-Independent Weight Perturbations on Mini-Batches
• Weight Uncertainty in Neural Networks
• Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning
• Variational Inference: A Review for Statisticians

Cited by

• Open-Set LiDAR Panoptic Segmentation Guided by Uncertainty-Aware Learning

Related

• Uncertainty aware audiovisual activity recognition using deep Bayesian variational inference
• Uncertainty-sensitive Activity Recognition: a Reliability Benchmark and the CARING Models
• Specifying Weight Priors in Bayesian Deep Neural Networks with Empirical Bayes
• Correlated Parameters to Accurately Measure Uncertainty in Deep Neural Networks
• BayesCap: Bayesian Identity Cap for Calibrated Uncertainty in Frozen Neural Networks