Deep Reinforcement Learning with Mixed Convolutional Network

TL;DR

This paper introduces a mixed convolutional neural network that combines sensor data and raw images to improve autonomous driving in a simulated racing environment, achieving superior performance over existing models.

Contribution

It presents a novel mixed CNN model that integrates sensor inputs with image data for enhanced driving performance in CarRacing-v0.

Findings

The mixed model outperforms AlexNet and VGG16 in average reward.

Data augmentation increased dataset size fourfold.

Sensor and image data integration improves boundary detection and driving accuracy.

Abstract

Recent research has shown that map raw pixels from a single front-facing camera directly to steering commands are surprisingly powerful. This paper presents a convolutional neural network (CNN) to playing the CarRacing-v0 using imitation learning in OpenAI Gym. The dataset is generated by playing the game manually in Gym and used a data augmentation method to expand the dataset to 4 times larger than before. Also, we read the true speed, four ABS sensors, steering wheel position, and gyroscope for each image and designed a mixed model by combining the sensor input and image input. After training, this model can automatically detect the boundaries of road features and drive the robot like a human. By comparing with AlexNet and VGG16 using the average reward in CarRacing-v0, our model wins the maximum overall system performance.