Memory-based Exploration-value Evaluation Model for Visual Navigation

We propose a hierarchical visual navigation solution, called Memory-based Exploration-value Evaluation Model (MEEM), to improve the agent's navigation performance. MEEM employs a hierarchical policy to tackle the challenge of sparse rewards, holds an episodic memory to store the historical information of the agent, and applies an Exploration-value Evaluation Model to calculate an exploration-value for action planning at each location in the observable area. We experimentally verify MEEM by navigation performance comparison on two datasets including the grid-map dataset and the 3D scenes Gibson dataset, where our approach achieves state-of-the-art performance on both. Specifically, the overall success rate of MEEM is 95% on the grid-map dataset while the best competitor reaches 68% only. As for the Gibson dataset, the success rate of ours and the best competitor SemExp are 69. 8% and 54. 4%, respectively. Ablation analysis on the tile-map dataset indicates that all three components of MEEM have positive effects.

Authors

• Yongquan Feng
• Liyang Xu
• Minglong Li
• Ruochun Jin
• Da Huang
• Shaowu Yang
• Wenjing Yang 0002

Keywords

• Visualization
• Three-dimensional displays
• Automation
• Navigation
• Semantic segmentation
• Planning
• Machine Vision
• Action Plan
• Episodic Memory
• Historical Information
• 3D Scene
• Performance Of Agents
• Navigation Performance
• Spatial Information
• State Space
• Upper Layer
• Adam Optimizer
• Current Position
• Lower Layer
• Visual Task
• Temporal Information
• Optimal Policy
• Intelligence Agencies
• Size Of Map
• Grid Map
• Simultaneous Localization And Mapping
• Semantic Map
• Position Of Agent
• Memory Structure
• Training Policy
• Environment Map
• Value Iteration
• Real-world Environments