DOI QR코드

DOI QR Code

Out-layer를 제거한 End to End 자율주행 시스템

End to End Autonomous Driving System using Out-layer Removal

  • 정승혁 (전남대학교 ICT융합시스템공학과 ) ;
  • 윤동호 (한국생산기술연구원 ) ;
  • 홍성훈 (전남대학교 ICT융합시스템공학과)
  • Seung-Hyeok Jeong (Department of ICT Convergence System Engineering, Chonnam University ) ;
  • Dong-Ho Yun (Korea Institute of Industrial Technology ) ;
  • Sung-Hun Hong (Department of ICT Convergence System Engineering, Chonnam University)
  • 투고 : 2022.12.21
  • 심사 : 2023.01.23
  • 발행 : 2023.02.28

초록

본 논문에서는 비전 센서 기반 시스템의 차선 이탈과 신호등 오인식 등을 개선하기 위해 End to End 모델을 활용한 자율주행 시스템을 제안한다. End to End 학습은 다양한 환경 조건에 대해 확장을 할 수 있다. 비전 센서 기반 모형 자동차를 이용하여 주행 데이터를 수집한다. 수집한 데이터를 이용하여 기존의 데이터와 아웃레이어를 제거한 데이터로 구성한다. 입력 데이터인 카메라 이미지 데이터, 출력 데이터인 속도와 조향 데이터로 클래스를 구성하고 End to End 모델을 활용하여 데이터 학습을 수행하였다. 학습된 모델의 신뢰성을 확인했다. 모형 자동차에 학습한 End to End 모델을 적용하여 이미지 데이터로 조향각을 예측한다. 모형 자동차의 학습 결과, 아웃레이어를 제거한 모델이 기존 모델보다 향상된 것을 볼 수 있다.

In this paper, we propose an autonomous driving system using an end-to-end model to improve lane departure and misrecognition of traffic lights in a vision sensor-based system. End-to-end learning can be extended to a variety of environmental conditions. Driving data is collected using a model car based on a vision sensor. Using the collected data, it is composed of existing data and data with outlayers removed. A class was formed with camera image data as input data and speed and steering data as output data, and data learning was performed using an end-to-end model. The reliability of the trained model was verified. Apply the learned end-to-end model to the model car to predict the steering angle with image data. As a result of the learning of the model car, it can be seen that the model with the outlayer removed is improved than the existing model.

키워드

과제정보

본 논문은 2023년도 한국생산기술연구원 기관주요사업 "스마트 모빌리티 핵심 요소기술 개발(1/3))(kitech JA-23-0011)" 의 지원을 받아 수행된 것임

참고문헌

  1. Y. LeCun, Y. Bengio, and G. Hinton, "Deep learning," Nature, Vol.521, No.7553, pp.436-444, 2015. https://doi.org/10.1038/nature14539
  2. A. Krizhevsky, I. Sutskever, and G. E. Hinton, "ImageNet classification with deep convolutional neural networks," Communications of the ACM, Vol 60, No.6, pp.84-90, 2017. https://doi.org/10.1145/3065386
  3. http://www.image-net.org/challenges/LSVRC
  4. O. Russakovsky et al., "ImageNet Large Scale Visual Recognition Challenge," International Journal of Computer Vision, Vol.115, No.3, pp.211-252, 2015. https://doi.org/10.1007/s11263-015-0816-y
  5. Rashmi, M. Kumar, and R. Saxena, "Algorithm and Technique on Various Edge Detection : A Survey," Signal & Image Processing An International Journal, Vol.4, No.3, pp.65-75, 2013. https://doi.org/10.5121/sipij.2013.4306
  6. P. Mukhopadhyay and B. B. Chaudhuri, "A survey of Hough Transform," Pattern Recognition, Vol.48, No.3, pp.993-1010, 2015. https://doi.org/10.1016/j.patcog.2014.08.027
  7. https://blog.naver.com/autoa2z/222069303400
  8. C. Chen, A. Seff, A. Kornhauser, and J. Xiao, "DeepDriving: Learning Affordance for Direct Perception in Autonomous Driving," 2015 IEEE International Conference on Computer Vision (ICCV). IEEE, 2015.
  9. Pomerleau, D. A. "Alvinn: An autonomous land vehicle in a neural network." Advances in neural information processing systems, 1. 1988.
  10. Y. LeCun, U. Muller, J. Ben, E. Cosatto, and B. Flepp. "Off-road obstacle avoidance through end-to-end learning," In Proceedings of the 18th International Conference on Neural Information Processing Systems (NIPS'05), pp.739-746, 2015.
  11. H. Xu, Y. Gao, F. Yu, and T. Darrell, "End-to-End Learning of Driving Models from Large-Scale Video Datasets," 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 2017.
  12. S. Baluja, "Evolution of an artificial neural network based autonomous land vehicle controller," IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics), Vol.26, No.3, Institute of Electrical and Electronics Engineers (IEEE), pp.450-463, 1996. https://doi.org/10.1109/3477.499795
  13. J. Koutnik, G. Cuccu, J. Schmidhuber, and F. Gomez, "Evolving large-scale neural networks for vision-based reinforcement learning," Proceedings of the 15th annual conference on Genetic and evolutionary computation. ACM, 2013.
  14. A. E. Sallab, M. Abdou, E. Perot, and S. Yogamani, "Deep Reinforcement Learning framework for Autonomous Driving," Electronic Imaging, Vol.29, No.19, pp.70-76, 2017. https://doi.org/10.2352/ISSN.2470-1173.2017.19.AVM-023
  15. M. Bojarski et al., "End to End Learning for Self-Driving Cars." arXiv, 2016.
  16. A. Kendall et al., "Learning to Drive in a Day," 2019 International Conference on Robotics and Automation (ICRA). IEEE, 2019.