Preprocessing Techniques Applied in Imitation Learning for Autonomous Vehicles

• Tatianna Aviz UFPA
• Wellington Lobato UNICAMP
• Denis Rosário UFPA
• Eduardo Cerqueira UFPA

DOI: https://doi.org/10.5753/wperformance.2024.3095

Abstract

Autonomous driving systems often use Conditional Imitation Learning (CIL) to learn human driving behavior. In CIL, a dataset with demonstrations from a driver is used to train a model that replicates his driving behavior. However, the generalization ability of the model tends to be reduced in unfamiliar driving scenarios. Enhancing the diversity of training data via preprocessing approaches may assist in improving the capacity of the model to generalize. Therefore, this paper proposes an evaluation of data pre-processing techniques for training models based on CIL. The results demonstrate that, together, data augmentation and normalization techniques increase the generalization capacity of CIL, obtaining a 62.81% better value compared to other approaches.

References

Codevilla, F., Lopez, A. M., Koltun, V., and Dosovitskiy, A. (2018a). On offline evaluation of vision-based driving models. In Proceedings of the European Conference on Computer Vision (ECCV), pages 236–251.

Codevilla, F., Müller, M., López, A., Koltun, V., and Dosovitskiy, A. (2018b). End-to-end driving via conditional imitation learning. In IEEE International Conference on Robotics and Automation (ICRA).

Codevilla, F., Santana, E., Lopez, A. M., and Gaidon, A. (2019). Exploring the limitations of behavior cloning for autonomous driving. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV).

Dosovitskiy, A., Ros, G., Codevilla, F., Lopez, A., and Koltun, V. (2017). CARLA: An open urban driving simulator. In Proceedings of the 1st Annual Conference on Robot Learning.

Eraqi, H. M., Moustafa, M. N., and Honer, J. (2020). Efficient occupancy grid mapping and camera-lidar fusion for conditional imitation learning driving. In IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC).

Eraqi, H. M., Moustafa, M. N., and Honer, J. (2022). Dynamic conditional imitation learning for autonomous driving. IEEE Transactions on Intelligent Transportation Systems, 23(12):22988–23001.

Hawke, J., Shen, R., Gurau, C., Sharma, S., Reda, D., Nikolov, N., Mazur, P., Micklethwaite, S., Griffiths, N., Shah, A., and Kndall, A. (2020). Urban driving with conditional imitation learning. In 2020 IEEE International Conference on Robotics and Automation (ICRA).

Hu, A., Corrado, G., Griffiths, N., Murez, Z., Gurau, C., Yeo, H., Kendall, A., Cipolla, R., and Shotton, J. (2022). Model-based imitation learning for urban driving. In Advances in Neural Information Processing Systems, volume 35, pages 20703–20716. Curran Associates, Inc.

Junior, W. L., de Souza, A. M., Cerqueira, E., Rosário, D., and Villas, L. (2021). Mecanismo eficiente de localização cooperativa para veículos autônomos conectados. In Anais do XXXIX Simpósio Brasileiro de Redes de Computadores e Sistemas Distribuídos, pages 336–349. SBC.

Le Mero, L., Yi, D., Dianati, M., and Mouzakitis, A. (2022). A survey on imitation learning techniques for end-to-end autonomous vehicles. IEEE Transactions on Intelligent Transportation Systems, 23(9):14128–14147.

Ly, A. O. and Akhloufi, M. (2021). Learning to drive by imitation: An overview of deep behavior cloning methods. IEEE Transactions on Intelligent Vehicles, 6(2):195–209.

Maharana, K., Mondal, S., and Nemade, B. (2022). A review: Data pre-processing and data augmentation techniques. Global Transitions Proceedings, 3(1):91–99.

Teng, S., Chen, L., Ai, Y., Zhou, Y., Xuanyuan, Z., and Hu, X. (2023). Hierarchical interpretable imitation learning for end-to-end autonomous driving. IEEE Transactions on Intelligent Vehicles, 8(1):673–683.

Xiao, Y., Codevilla, F., Gurram, A., Urfalioglu, O., and López, A. M. (2022). Multimodal end-to-end autonomous driving. IEEE Transactions on Intelligent Transportation Systems, 23(1):537–547.

Zhang, E., Zhou, H., Ding, Y., Zhao, J., and Ye, C. (2020). Learning how to avoiding obstacles for end-to-end driving with conditional imitation learning. In Proceedings of the 2019 2nd International Conference on Signal Processing and Machine Learning. Association for Computing Machinery.

Zheng, B., Verma, S., Zhou, J., Tsang, I. W., and Chen, F. (2022). Imitation learning: Progress, taxonomies and challenges. IEEE Transactions on Neural Networks and Learning Systems, pages 1–16.