CNN Based Traffic Sign Recognition for Mini Autonomous Vehicles

Advanced driving assistance systems (ADAS) could perform basic object detection and classification to alert drivers for road conditions, vehicle speed regulation, and etc. With the advances in the new hardware and software platforms, deep learning has been used in ADAS technologies. Traffic signs are an important part of road infrastructure. So, it is very important task to detect and classify traffic signs for autonomous vehicles. In this paper, we firstly create a traffic sign dataset from ZED stereo camera mounted on the top of Racecar mini autonomous vehicle and we use Tiny-YOLO real-time object detection and classification system to detect and classify traffic signs. Then, we test the model on our dataset in terms of accuracy, loss, precision and intersection over union performance metrics.

[1]  Sertac Karaman,et al.  Project-based, collaborative, algorithmic robotics for high school students: Programming self-driving race cars at MIT , 2017, 2017 IEEE Integrated STEM Education Conference (ISEC).

[2]  Mykel J. Kochenderfer,et al.  Imitating driver behavior with generative adversarial networks , 2017, 2017 IEEE Intelligent Vehicles Symposium (IV).

[3]  Kyunghyun Cho,et al.  Query-Efficient Imitation Learning for End-to-End Simulated Driving , 2017, AAAI.

[4]  Kang-Hyun Jo,et al.  A comparative study of classification methods for traffic signs recognition , 2014, 2014 IEEE International Conference on Industrial Technology (ICIT).

[5]  Kang-Hyun Jo,et al.  Graph-based Approach for Robust Road Guidance Sign Recognition from Differently Exposed Images , 2009, J. Univers. Comput. Sci..

[6]  Yadong Mu,et al.  Learning End-to-End Autonomous Steering Model from Spatial and Temporal Visual Cues , 2017, VSCC '17.

[7]  Geoffrey E. Hinton,et al.  Deep Learning , 2015, Nature.

[8]  Mohamed El Ansari,et al.  Traffic sign detection and recognition based on random forests , 2016, Appl. Soft Comput..

[9]  Yadong Mu,et al.  Deep Steering: Learning End-to-End Driving Model from Spatial and Temporal Visual Cues , 2017, ArXiv.

[10]  O. Yalcin,et al.  Detection of road boundaries and obstacles using LIDAR , 2014, 2014 6th Computer Science and Electronic Engineering Conference (CEEC).

[11]  Ram Gopal Raj,et al.  Real-Time (Vision-Based) Road Sign Recognition Using an Artificial Neural Network , 2017, Sensors.

[12]  Jürgen Schmidhuber,et al.  Evolving large-scale neural networks for vision-based TORCS , 2013, FDG.

[13]  Ahmet Sayar,et al.  Approaches of Road Boundary and Obstacle Detection Using LIDAR , 2013 .

[14]  Ahmet Sayar,et al.  End-to-end learning model design for steering autonomous vehicle , 2018, 2018 26th Signal Processing and Communications Applications Conference (SIU).

[15]  Yang Gao,et al.  End-to-End Learning of Driving Models from Large-Scale Video Datasets , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[16]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[17]  Cüneyt Güzelis,et al.  Object recognition and detection with deep learning for autonomous driving applications , 2017, Simul..

[18]  Alex Graves,et al.  Asynchronous Methods for Deep Reinforcement Learning , 2016, ICML.

[19]  Jürgen Schmidhuber,et al.  Evolving large-scale neural networks for vision-based reinforcement learning , 2013, GECCO '13.

[20]  Ali Farhadi,et al.  YOLO9000: Better, Faster, Stronger , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[21]  Geoffrey J. Gordon,et al.  A Reduction of Imitation Learning and Structured Prediction to No-Regret Online Learning , 2010, AISTATS.

[22]  Anjan Gudigar,et al.  Kernel Based Automatic Traffic Sign Detection and Recognition Using SVM , 2012, ICECCS 2012.

[23]  Yan Han,et al.  Robust traffic sign recognition with feature extraction and k-NN classification methods , 2015, 2015 IEEE International Conference on Electro/Information Technology (EIT).