Object Detection for Autonomous Vehicles with Sensor-based Technology Using YOLO

The year 2020 has been a tough year with the global pandemic situation, and the utmost priority is to live in a clean, green, and safe environment. One of the areas that the governments are emphasizing for the readiness of our ecosystem is autonomous and contactless environments in adapting to the new norm. Thus, Autonomous Vehicle (AV) is a promising technology to bring forward. One of the critical aspects of Autonomous Navigation is object detection. Most AV use multiple sensors to detect objects, such as cameras, radar and Light Detection and Ranging sensor (LiDAR). Nowadays, the LiDAR sensor is widely implemented due to the ability to detect objects in the form of pulsed lasers, benefiting in low-light object detection. However, even with advanced technology, poor programming can affect the performance of object detection system. Thus, the study explores the state-of-the-art of You Only Look Once (YOLO) algorithms namely Tiny-YOLO and Complex-YOLO for object detection on KITTI dataset. Their performances were compared based on accuracy, precision, and recall metrics. The results showed that the Complex-YOLO has better performance as the mean average precision is higher than the Tiny-YOLO model when tested with equal parameters.

PDF

___

[1] Yahya, M. A., Abdul-Rahman, S., & Mutalib, S. (2020). Object detection for autonomous vehicle with Lidar using deep learning. 2020 IEEE 10th International Conference on System Engineering and Technology, ICSET 2020 - Proceedings, 207–212. https://doi.org/10.1109/ICSET51301.2020.9265358
[2] NTSB. (2018). Preliminary Report HWY18MH010. 4. https://www.ntsb.gov/investigations/AccidentReports/Reports/HW Y18MH010-prelim.pdf
[3] Combs, T. S., Sandt, L. S., Clamann, M. P., & McDonald, N. C. (2019). Automated Vehicles and Pedestrian Safety: Exploring the Promise and Limits of Pedestrian Detection. American Journal of Preventive Medicine, 56(1), 1–7. https://doi.org/10.1016/j.amepre.2018.06.024
[4] Roth, M., Jargot, D., & Gavrila, D. M. (2019). Deep End-to-end 3D Person Detection from Camera and Lidar. 2019 IEEE Intelligent Transportation Systems Conference, ITSC 2019, 521–527. https://doi.org/10.1109/ITSC.2019.8917366
[5] Srivastava, S., Divekar, A. V., Anilkumar, C., Naik, I., Kulkarni, V., & Pattabiraman, V. (2021). Comparative analysis of deep learning image detection algorithms. Journal of Big Data, 8(1). https://doi.org/10.1186/s40537-021-00434-w
[6] Mugunthan, N., Sb, B., Harini, C., Naresh, V. H., & V, P. V. (2020). Comparison Review on LiDAR vs Camera in Autonomous Vehicle. IRJET, 07(08), 4242–4246.
[7] Review, A. (2019). Lidar vs. Cameras for Self Driving Cars – What’s Best? 1–7. https://www.autopilotreview.com/lidar-vscameras- self-driving-cars/
[8] Royo, S., & Ballesta-Garcia, M. (2019). An overview of lidar imaging systems for autonomous vehicles. Applied Sciences (Switzerland), 9(19). https://doi.org/10.3390/app9194093
[9] Borcs, A., Nagy, B., & Benedek, C. (2017). Instant Object Detection in Lidar Point Clouds. IEEE Geoscience and Remote Sensing Letters, 14(7), 992–996. https://doi.org/10.1109/lgrs.2017.2674799
[10] Bagloee, S. A., Tavana, M., Asadi, M., & Oliver, T. (2016). Autonomous vehicles: challenges, opportunities, and future implications for transportation policies. Journal of Modern Transportation, 24(4), 284–303. https://doi.org/10.1007/s40534- 016-0117-3
[11] Engels, G., Aranjuelo, N., Arganda-Carreras, I., Nieto, M., & Otaegui, O. (2020). 3D object detection from LiDAR data using distance dependent feature extraction. VEHITS 2020 - Proceedings of the 6th International Conference on Vehicle Technology and Intelligent Transport Systems, March, 289–300. https://doi.org/10.5220/0009330402890300
[12] Bandyopadhyay, H. (n.d.). YOLO: Real-Time Object Detection Explained. Retrieved August 9, 2021, from https://www.v7labs.com/blog/yolo-object-detection#two-stagedetectors
[13] Simon, M., Milz, S., Amende, K., & Gross, H.-M. (2018). Complex- YOLO: Real-time 3D Object Detection on Point Clouds. 1–14. http://arxiv.org/abs/1803.06199
[14] Choi, J., Chun, D., Kim, H., & Lee, H. J. (2019). Gaussian YOLOv3: An accurate and fast object detector using localization uncertainty for autonomous driving. Proceedings of the IEEE International Conference on Computer Vision, 2019-October, 502–511. https://doi.org/10.1109/ICCV.2019.00059
[15] Yin, X., Sasaki, Y., Wang, W., & Shimizu, K. (2020). 3D Object Detection Method Based on YOLO and K-Means for Image and Point Clouds. ArXiv. http://arxiv.org/abs/2005.02132
[16] Adarsh, Pranav & Rathi, Pratibha & Kumar, Manoj. (2020). YOLO v3-Tiny: Object Detection and Recognition using one stage improved model. 687-694. 10.1109/ICACCS48705.2020.9074315.
[17] Benjdira, B., Khursheed, T., Koubaa, A., Ammar, A., & Ouni, K. (2018). Car Detection using Unmanned Aerial Vehicles: Comparison between Faster R-CNN and YOLOv3. arXiv. https://arxiv.org/abs/1812.10968
[18] Yi, Z., Yongliang, S., & Jun, Z. (2019). An improved tiny-yolov3 pedestrian detection algorithm. Optik, 183, 17–23. https://doi.org/10.1016/j.ijleo.2019.02.038
[19] Gong, X., Ma, L., & Ouyang, H. (2020). An improved method of Tiny YOLOV3. IOP Conference Series: Earth and Environmental Science, 440(5), 052025. https://doi.org/10.1088/1755- 1315/440/5/052025
[20] Fang, W., Wang, L., Ren, P., 2020. Tinier-YOLO: A Real-Time Object Detection Method for Constrained Environments. IEEE Access 8, 1935–1944.. doi:10.1109/access.2019.2961959
[21] Geiger, A., Lenz, P., Stiller, C., & Urtasun, R. (2013). Vision meets robotics: The KITTI dataset. The International Journal of Robotics Research. The International Journal of Robotics Research, October, 1–6.
[22] Deepak Ghimire. (2019, December 11). Complex-YOLOv3: PyTorch implementation of Complex-YOLO paper with YoloV3. https://github.com/ghimiredhikura/Complex-YOLOv3