Towards Monocular Vision based Obstacle Avoidance through Deep Reinforcement Learning

Abstract

Obstacle avoidance is a fundamental requirement for autonomous robots which operate in, and interact with, the real world. When perception is limited to monocular vision avoiding collision becomes significantly more challenging due to the lack of 3D information. Conventional path planners for obstacle avoidance require tuning a number of parameters and do not have the ability to directly benefit from large datasets and continuous use. In this paper, a dueling architecture based deep double-Q network (D3QN) is proposed for obstacle avoidance, using only monocular RGB vision. Based on the dueling and double-Q mechanisms, D3QN can efficiently learn how to avoid obstacles in a simulator even with very noisy depth information predicted from RGB image. Extensive experiments show that D3QN enables twofold acceleration on learning compared with a normal deep Q network and the models trained solely in virtual environments can be directly transferred to real robots, generalizing well to various new environments with previously unseen dynamic objects.

Accepted by RSS 2017 workshop New Frontiers for Deep Learning in Robotics

References in corpus (9)

• Adam: A Method for Stochastic Optimization
• Dueling Network Architectures for Deep Reinforcement Learning
• DeepVO: Towards End-to-End Visual Odometry with Deep Recurrent Convolutional Neural Networks
• Learning to Navigate in Complex Environments
• Deep Neural Network for Real-Time Autonomous Indoor Navigation
• Towards Cognitive Exploration through Deep Reinforcement Learning for Mobile Robots
• Obstacle Avoidance through Deep Networks based Intermediate Perception
• VidLoc: A Deep Spatio-Temporal Model for 6-DoF Video-Clip Relocalization
• Learning to Fly by Crashing

Cited by in corpus (14)

• Perception and Navigation in Autonomous Systems in the Era of Learning: A Survey
• Goal-Guided Transformer-Enabled Reinforcement Learning for Efficient Autonomous Navigation
• A real-time dynamic obstacle tracking and mapping system for UAV navigation and collision avoidance with an RGB-D camera
• Realtime Collision Avoidance for Mobile Robots in Dense Crowds using Implicit Multi-sensor Fusion and Deep Reinforcement Learning
• Learning to Sequence Robot Behaviors for Visual Navigation
• Learning with Stochastic Guidance for Navigation
• Danger-aware Adaptive Composition of DRL Agents for Self-navigation
• Improved Reinforcement Learning through Imitation Learning Pretraining Towards Image-based Autonomous Driving
• Energy-Efficient Parking Analytics System using Deep Reinforcement Learning
• OF-VO: Efficient Navigation among Pedestrians Using Commodity Sensors
• End-to-end Decentralized Multi-robot Navigation in Unknown Complex Environments via Deep Reinforcement Learning
• When Autonomous Systems Meet Accuracy and Transferability through AI: A Survey
• Learning with Training Wheels: Speeding up Training with a Simple Controller for Deep Reinforcement Learning
• Autonomous quadrotor obstacle avoidance based on dueling double deep recurrent Q-learning with monocular vision