Object-sensitive Deep Reinforcement Learning

2018/09/17 by Yuezhang Li, Li, Yuezhang, Katia Sycara +3 · 3 citations
Computer Science · Engineering · Mathematics · #Advanced Neural Network Applications #Adversarial Robustness in Machine Learning #Artificial intelligence #Computer Vision and Pattern Recognition (cs.CV) #Computer science #Computer vision #Deep learning #Engineering #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Object (grammar) #Reinforcement #Reinforcement Learning in Robotics #Reinforcement learning #cs.CV #cs.LG #stat.ML

paper · pdf · doi:10.48550/arxiv.1809.06064

published in arXiv (Cornell University) (Cornell University) · 15 pages, 6 figures, Accepted at 3rd Global Conference on Artificial Intelligence (GCAI-17), Miami, 2017

arxiv created 2018/09/17 · openalex publication_date 2018/09/17 · arxiv updated 2018/09/18 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05

Abstract

Deep reinforcement learning has become popular over recent years, showing superiority on different visual-input tasks such as playing Atari games and robot navigation. Although objects are important image elements, few work considers enhancing deep reinforcement learning with object characteristics. In this paper, we propose a novel method that can incorporate object recognition processing to deep reinforcement learning models. This approach can be adapted to any existing deep reinforcement learning frameworks. State-of-the-art results are shown in experiments on Atari games. We also propose a new approach called "object saliency maps" to visually explain the actions made by deep reinforcement learning agents.

Citations

• Improving neural networks by preventing co-adaptation of feature detectors
• On learning to localize objects with minimal supervision
• Deep Fragment Embeddings for Bidirectional Image Sentence Mapping
• Human-level control through deep reinforcement learning
• Backpropagation Applied to Handwritten Zip Code Recognition
• Distinctive Image Features from Scale-Invariant Keypoints
• Dueling Network Architectures for Deep Reinforcement Learning
• Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps

Cited by

• DRTA: Dynamic Reward Scaling for Reinforcement Learning in Time Series Anomaly Detection
• Enhancing Learning Path Recommendation via Multi-task Learning
• Anomaly Detection in Time Series Data Using Reinforcement Learning, Variational Autoencoder, and Active Learning

Related

• Transparency and Explanation in Deep Reinforcement Learning Neural Networks
• Relevance-Guided Modeling of Object Dynamics for Reinforcement Learning
• Investigating Simple Object Representations in Model-Free Deep Reinforcement Learning
• OCAtari: Object-Centric Atari 2600 Reinforcement Learning Environments
• Unsupervised Video Object Segmentation for Deep Reinforcement Learning