IEMEK Journal of Embedded Systems and Applications (대한임베디드공학회논문지)

Institute of Embedded Engineering of Korea (대한임베디드공학회)
권호 표지
DOI QR코드

DOI QR Code

Object-aware Depth Estimation for Developing Collision Avoidance System

객체 영역에 특화된 뎁스 추정 기반의 충돌방지 기술개발

  • Received : 2023.12.28
  • Accepted : 2024.02.14
  • Published : 2024.04.30
Download PDF
⟨ Previous Next ⟩

Abstract

Collision avoidance system is important to improve the robustness and functional safety of autonomous vehicles. This paper proposes an object-level distance estimation method to develop a collision avoidance system, and it is applied to golfcarts utilized in country club environments. To improve the detection accuracy, we continually trained an object detection model based on pseudo labels generated by a pre-trained detector. Moreover, we propose object-aware depth estimation (OADE) method which trains a depth model focusing on object regions. In the OADE algorithm, we generated dense depth information for object regions by utilizing detection results and sparse LiDAR points, and it is referred to as object-aware LiDAR projection (OALP). By using the OALP maps, a depth estimation model was trained by backpropagating more gradients of the loss on object regions. Experiments were conducted on our custom dataset, which was collected for the travel distance of 22 km on 54 holes in three country clubs under various weather conditions. The precision and recall rate were respectively improved from 70.5% and 49.1% to 95.3% and 92.1% after the continual learning with pseudo labels. Moreover, the OADE algorithm reduces the absolute relative error from 4.76% to 4.27% for estimating distances to obstacles.

Keywords

Acknowledgement

본 과제 (결과물)는 교육부와 한국연구재단의 재원으로 지원을 받아 수행된 사회맞춤형 산학협력 선도대학 (LINC+) 육성사업의 연구결과입니다. 이 연구는 2023년도 산업통상자원부 및 산업기술평가관리원 (KEIT) 연구비 지원에 의한 연구임 (20023305).

References

  1. H. Kang, "고령화 사회와 환대산업의 전망," Journal of Korea Entertainment Industry Association, Vol, 10, No. 1, pp. 10-16, 2018. 
  2. M. Kim, S. Kang, "Demand Prediction Analysis of the Domestic Golf Courses Applying Exponential Smoothing Model," Korean Journal of Convergence Science (KJCS), Vol. 11, No. 11, pp. 113-131, 2022. 
  3. J. Kim, S. Ye, Y. Kang, S. Han, "Design and Implementation of Android Based Golf Cart Autonomous Driving Simulation System," Proceedings of the Korea Information Processing Society Conference, Vol. 19, No. 1, pp. 843-845, 2012. 
  4. J. Choi, I. Yeo, J. Jo, "Autonomous Traveling of Unmanned Golf-Car using GPS and Vision System," Journal of the Korean Society for Precision Engineering, Vol.. 26, No. 6, pp. 74-80, 2009. 
  5. H. Park, J. Lee, K. Lee, I, Choi, "Development of Cost-effective Self-Driving Module for Golf Cart," Proceddings of Institude of Control, Robotics and Systems, pp. 208-209, 2020. 
  6. J. Kim, G. Son, Y. Jung, "Yolo v3-based Golf Course Environmental Object Detection Method," The Journal of Korean Institute of Communications and Information Sciences, pp. 1164-1165, 2020. 
  7. S. Ren, K. He, R. Girshick, J. Sun, "Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks," Proceedings of Advances in Neural Information Processing Systems, Vol. 28, 2015. 
  8. T. Y. Lin, P. Dollar, R. Girshick, K. He, B. Hariharan, S. Belongie, "Feature Pyramid Networks for Object Detection," Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 2117-2125, 2017. 
  9. N. Carion, F. Massa, G. Synnaeve, N. Usunier, A.Kirillov, S. Zagoruyko, "End-to-End Object Detection with Transformers," European Conference on Computer Vision (ECCV), pp. 216-229, 2020. 
  10. J. Lee, M. Han, D. Ko, I. Suh, "From Big to Small: Multi-Scale Local Planar Guidance for Monocular Depth Estimation," ArXiv:1907.10326, 2019. 
  11. S. Bhat, I. Alhashim, P. Wonka, "AdaBins: Depth Estimation Using Adaptive Bins," Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 4009-4018, 2021. 
  12. Z. Li, X. Wang, X. Liu, J. Jiang, "BinsFormer: Revisiting Adaptive Bins for Monocular Depth Estimation," ArXiv:2204.009872022, 2022. 
  13. 송지민, 이상준, "비지도학습 기반의 뎁스 추정을 위한 지식 증류 기법," 대한임베디드공학회논문지, 제17권, 제4호, pp. 209-215, 2022.  https://doi.org/10.14372/IEMEK.2022.17.4.209
  14. D. Tsai, S. Worrall, M. Shan, A. Lohr, E. Nebot, "Optimising the Selection of Samples for Robust Lidar Camera Calibration," Proceddings of 2021 IEEE International Intelligent Transportation Systems Conference (ITSC), pp. 2631-2638, 2021. 
  15. J. L. Bentley, "Multidimensional Binary Search Trees Used for Associative Searching," Communications of the ACM, Vol. 18, pp. 509-517, 1975.  https://doi.org/10.1145/361002.361007
  16. J. Redmon, S, Divvala, R, Girshick, A. Farhadi, "You Only Look Once: Unified, Real-Time Object Detection," Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 779-788, 2016. 
  17. C. Wang, H. Liao, Y. Wu, P. Chen, J. Hsieh, I. Yeh, "CSPNet: A New Backbone that can Enhance Learning Capability of CNN," Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, pp. 390-391, 2020. 
  18. S. Liu, L. Qi, H. Qin, J. Shi, J. JiaPath, "Aggregation Network for Instance Segmentation," Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 8759-8768, 2018. 
  19. T. Lin, M. Maire, S. Belongie, J. Hays, P. Perona, D. Ramanan, P. Dollar, C. Zitnick, "Microsoft COCO: Common Objects in Context," Proceedings of European Conference on Computer Vision (ECCV), pp. 740-755, 2014. 
  20. M. Yang, K. Yu, C. Zhang, Z. Li, K. Yang, "DenseASPP for Semantic Segmentation in Street Scenes," Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 3684-3692, 2018. 
  21. G. Huang, Z. Liu, L. Maaten, K. Weinberger, "Densely Connected Convolutional Networks," Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 4700-4708, 2017. 
  22. A. Geiger, P. Lenz, R. Urtasun, "Are we Ready for Autonomous Driving? The KITTI Vision Benchmark Suite," Proceedings of 2012 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 3354-3361, 2012.