Variational Autoencoder for End-to-End Control of Autonomous Driving with Novelty Detection and Training De-biasing
This paper introduces a new method for end-to-end training of deep neural networks (DNNs) and evaluates it in the context of autonomous driving. DNN training has been shown to result in high accuracy for perception to action learning given sufficient training data. However, the trained models may fa...
| Main Authors: | , , , , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | en_US |
| Published: |
2018
|
| Online Access: | http://hdl.handle.net/1721.1/118139 https://orcid.org/0000-0002-9334-1706 https://orcid.org/0000-0002-2225-7275 https://orcid.org/0000-0001-5473-3566 |
| _version_ | 1826191523092365312 |
|---|---|
| author | Amini, Alexander Araki, Brandon Rus, Daniela Schwarting, Wilko Rosman, Guy Karaman, Sertac Rus, Daniela L |
| author2 | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory |
| author_facet | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Amini, Alexander Araki, Brandon Rus, Daniela Schwarting, Wilko Rosman, Guy Karaman, Sertac Rus, Daniela L |
| author_sort | Amini, Alexander |
| collection | MIT |
| description | This paper introduces a new method for end-to-end training of deep neural networks (DNNs) and evaluates it in the context of autonomous driving. DNN training has been shown to result in high accuracy for perception to action learning given sufficient training data. However, the trained models may fail without warning in situations with insufficient or biased training data. In this paper, we propose and evaluate a novel architecture for self-supervised learning of latent variables to detect the insufficiently trained situations. Our method also addresses training data imbalance, by learning a set of underlying latent variables that characterize the training data and evaluate potential biases. We show how these latent distributions can be leveraged to adapt and accelerate the training pipeline by training on only a fraction of the total dataset. We evaluate our approach on a challenging dataset for driving. The data is collected from a full-scale autonomous vehicle. Our method provides qualitative explanation for the latent variables learned in the model. Finally, we show how our model can be additionally trained as an end-to-end controller, directly outputting a steering control command for an autonomous vehicle. |
| first_indexed | 2024-09-23T08:57:29Z |
| format | Article |
| id | mit-1721.1/118139 |
| institution | Massachusetts Institute of Technology |
| language | en_US |
| last_indexed | 2024-09-23T08:57:29Z |
| publishDate | 2018 |
| record_format | dspace |
| spelling | mit-1721.1/1181392022-09-26T09:27:30Z Variational Autoencoder for End-to-End Control of Autonomous Driving with Novelty Detection and Training De-biasing Amini, Alexander Araki, Brandon Rus, Daniela Schwarting, Wilko Rosman, Guy Karaman, Sertac Rus, Daniela L Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Massachusetts Institute of Technology. Department of Aeronautics and Astronautics Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Massachusetts Institute of Technology. Department of Mechanical Engineering Amini, Alexander Schwarting, Wilko Rosman, Guy Karaman, Sertac Rus, Daniela L This paper introduces a new method for end-to-end training of deep neural networks (DNNs) and evaluates it in the context of autonomous driving. DNN training has been shown to result in high accuracy for perception to action learning given sufficient training data. However, the trained models may fail without warning in situations with insufficient or biased training data. In this paper, we propose and evaluate a novel architecture for self-supervised learning of latent variables to detect the insufficiently trained situations. Our method also addresses training data imbalance, by learning a set of underlying latent variables that characterize the training data and evaluate potential biases. We show how these latent distributions can be leveraged to adapt and accelerate the training pipeline by training on only a fraction of the total dataset. We evaluate our approach on a challenging dataset for driving. The data is collected from a full-scale autonomous vehicle. Our method provides qualitative explanation for the latent variables learned in the model. Finally, we show how our model can be additionally trained as an end-to-end controller, directly outputting a steering control command for an autonomous vehicle. 2018-09-18T16:21:36Z 2018-09-18T16:21:36Z 2018-10 Article http://purl.org/eprint/type/ConferencePaper http://hdl.handle.net/1721.1/118139 Amini, Alexander, Wilko Schwarting, Guy Rosman, Brandon Araki. Sertac Karaman and Daniela Rus. "Variational Autoencoder for End-to-End Control of Autonomous Driving with Novelty Detection and Training De-biasing." 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (Palacio Municipal de Congresos, Madrid, Spain, Oct.1-5 2018) https://orcid.org/0000-0002-9334-1706 https://orcid.org/0000-0002-2225-7275 https://orcid.org/0000-0001-5473-3566 en_US https://www.iros2018.org/ 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) Creative Commons Attribution-Noncommercial-Share Alike http://creativecommons.org/licenses/by-nc-sa/4.0/ application/pdf Amini |
| spellingShingle | Amini, Alexander Araki, Brandon Rus, Daniela Schwarting, Wilko Rosman, Guy Karaman, Sertac Rus, Daniela L Variational Autoencoder for End-to-End Control of Autonomous Driving with Novelty Detection and Training De-biasing |
| title | Variational Autoencoder for End-to-End Control of Autonomous Driving with Novelty Detection and Training De-biasing |
| title_full | Variational Autoencoder for End-to-End Control of Autonomous Driving with Novelty Detection and Training De-biasing |
| title_fullStr | Variational Autoencoder for End-to-End Control of Autonomous Driving with Novelty Detection and Training De-biasing |
| title_full_unstemmed | Variational Autoencoder for End-to-End Control of Autonomous Driving with Novelty Detection and Training De-biasing |
| title_short | Variational Autoencoder for End-to-End Control of Autonomous Driving with Novelty Detection and Training De-biasing |
| title_sort | variational autoencoder for end to end control of autonomous driving with novelty detection and training de biasing |
| url | http://hdl.handle.net/1721.1/118139 https://orcid.org/0000-0002-9334-1706 https://orcid.org/0000-0002-2225-7275 https://orcid.org/0000-0001-5473-3566 |
| work_keys_str_mv | AT aminialexander variationalautoencoderforendtoendcontrolofautonomousdrivingwithnoveltydetectionandtrainingdebiasing AT arakibrandon variationalautoencoderforendtoendcontrolofautonomousdrivingwithnoveltydetectionandtrainingdebiasing AT rusdaniela variationalautoencoderforendtoendcontrolofautonomousdrivingwithnoveltydetectionandtrainingdebiasing AT schwartingwilko variationalautoencoderforendtoendcontrolofautonomousdrivingwithnoveltydetectionandtrainingdebiasing AT rosmanguy variationalautoencoderforendtoendcontrolofautonomousdrivingwithnoveltydetectionandtrainingdebiasing AT karamansertac variationalautoencoderforendtoendcontrolofautonomousdrivingwithnoveltydetectionandtrainingdebiasing AT rusdanielal variationalautoencoderforendtoendcontrolofautonomousdrivingwithnoveltydetectionandtrainingdebiasing |