Addressing appearance change in outdoor robotics with adversarial domain adaptation
Appearance changes due to weather and seasonal conditions represent a strong impediment to the robust implementation of machine learning systems in outdoor robotics. While supervised learning optimises a model for the training domain, it will deliver degraded performance in application domains that...
| Main Authors: | , , |
|---|---|
| Format: | Conference item |
| Published: |
Institute of Electrical and Electronics Engineers
2017
|
| _version_ | 1797087254360883200 |
|---|---|
| author | Wulfeier, M Bewley, A Posner, H |
| author_facet | Wulfeier, M Bewley, A Posner, H |
| author_sort | Wulfeier, M |
| collection | OXFORD |
| description | Appearance changes due to weather and seasonal conditions represent a strong impediment to the robust implementation of machine learning systems in outdoor robotics. While supervised learning optimises a model for the training domain, it will deliver degraded performance in application domains that underlie distributional shifts caused by these changes. Traditionally, this problem has been addressed via the collection of labelled data in multiple domains or by imposing priors on the type of shift between both domains. We frame the problem in the context of unsupervised domain adaptation and develop a framework for applying adversarial techniques to adapt popular, state-of-the-art network architectures with the additional objective to align features across domains. Moreover, as adversarial training is notoriously unstable, we first perform an extensive ablation study, adapting many techniques known to stabilise generative adversarial networks, and evaluate on a surrogate classification task with the same appearance change. The distilled insights are applied to the problem of free-space segmentation for motion planning in autonomous driving. |
| first_indexed | 2024-03-07T02:33:12Z |
| format | Conference item |
| id | oxford-uuid:a7ef3081-6912-4fab-a9b2-922f7fe8d733 |
| institution | University of Oxford |
| last_indexed | 2024-03-07T02:33:12Z |
| publishDate | 2017 |
| publisher | Institute of Electrical and Electronics Engineers |
| record_format | dspace |
| spelling | oxford-uuid:a7ef3081-6912-4fab-a9b2-922f7fe8d7332022-03-27T02:57:53ZAddressing appearance change in outdoor robotics with adversarial domain adaptationConference itemhttp://purl.org/coar/resource_type/c_5794uuid:a7ef3081-6912-4fab-a9b2-922f7fe8d733Symplectic Elements at OxfordInstitute of Electrical and Electronics Engineers2017Wulfeier, MBewley, APosner, HAppearance changes due to weather and seasonal conditions represent a strong impediment to the robust implementation of machine learning systems in outdoor robotics. While supervised learning optimises a model for the training domain, it will deliver degraded performance in application domains that underlie distributional shifts caused by these changes. Traditionally, this problem has been addressed via the collection of labelled data in multiple domains or by imposing priors on the type of shift between both domains. We frame the problem in the context of unsupervised domain adaptation and develop a framework for applying adversarial techniques to adapt popular, state-of-the-art network architectures with the additional objective to align features across domains. Moreover, as adversarial training is notoriously unstable, we first perform an extensive ablation study, adapting many techniques known to stabilise generative adversarial networks, and evaluate on a surrogate classification task with the same appearance change. The distilled insights are applied to the problem of free-space segmentation for motion planning in autonomous driving. |
| spellingShingle | Wulfeier, M Bewley, A Posner, H Addressing appearance change in outdoor robotics with adversarial domain adaptation |
| title | Addressing appearance change in outdoor robotics with adversarial domain adaptation |
| title_full | Addressing appearance change in outdoor robotics with adversarial domain adaptation |
| title_fullStr | Addressing appearance change in outdoor robotics with adversarial domain adaptation |
| title_full_unstemmed | Addressing appearance change in outdoor robotics with adversarial domain adaptation |
| title_short | Addressing appearance change in outdoor robotics with adversarial domain adaptation |
| title_sort | addressing appearance change in outdoor robotics with adversarial domain adaptation |
| work_keys_str_mv | AT wulfeierm addressingappearancechangeinoutdoorroboticswithadversarialdomainadaptation AT bewleya addressingappearancechangeinoutdoorroboticswithadversarialdomainadaptation AT posnerh addressingappearancechangeinoutdoorroboticswithadversarialdomainadaptation |