Adversarial training for adverse conditions: Robust metric localisation using appearance transfer
We present a method of improving visual place recognition and metric localisation under very strong appearance change. We learn an invertable generator that can transform the conditions of images, e.g. from day to night, summer to winter etc. This image transforming filter is explicitly designed to...
| Main Authors: | , , |
|---|---|
| Format: | Conference item |
| Language: | English |
| Published: |
IEEE
2018
|
| _version_ | 1826261878827909120 |
|---|---|
| author | Porav, H Maddern, W Newman, P |
| author_facet | Porav, H Maddern, W Newman, P |
| author_sort | Porav, H |
| collection | OXFORD |
| description | We present a method of improving visual place recognition and metric localisation under very strong appearance change. We learn an invertable generator that can transform the conditions of images, e.g. from day to night, summer to winter etc. This image transforming filter is explicitly designed to aid and abet feature-matching using a new loss based on SURF detector and dense descriptor maps. A network is trained to output synthetic images optimised for feature matching given only an input RGB image, and these generated images are used to localize the robot against a previously built map using traditional sparse matching approaches. We benchmark our results using multiple traversals of the Oxford RobotCar Dataset over a year-long period, using one traversal as a map and the other to localise. We show that this method significantly improves place recognition and localisation under changing and adverse conditions, while reducing the number of mapping runs needed to successfully achieve reliable localisation. |
| first_indexed | 2024-03-06T19:27:31Z |
| format | Conference item |
| id | oxford-uuid:1c48f1e1-3c4f-4fd4-b09e-be8ef78deadc |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-03-06T19:27:31Z |
| publishDate | 2018 |
| publisher | IEEE |
| record_format | dspace |
| spelling | oxford-uuid:1c48f1e1-3c4f-4fd4-b09e-be8ef78deadc2022-03-26T11:04:52ZAdversarial training for adverse conditions: Robust metric localisation using appearance transferConference itemhttp://purl.org/coar/resource_type/c_5794uuid:1c48f1e1-3c4f-4fd4-b09e-be8ef78deadcEnglishSymplectic ElementsIEEE2018Porav, HMaddern, WNewman, PWe present a method of improving visual place recognition and metric localisation under very strong appearance change. We learn an invertable generator that can transform the conditions of images, e.g. from day to night, summer to winter etc. This image transforming filter is explicitly designed to aid and abet feature-matching using a new loss based on SURF detector and dense descriptor maps. A network is trained to output synthetic images optimised for feature matching given only an input RGB image, and these generated images are used to localize the robot against a previously built map using traditional sparse matching approaches. We benchmark our results using multiple traversals of the Oxford RobotCar Dataset over a year-long period, using one traversal as a map and the other to localise. We show that this method significantly improves place recognition and localisation under changing and adverse conditions, while reducing the number of mapping runs needed to successfully achieve reliable localisation. |
| spellingShingle | Porav, H Maddern, W Newman, P Adversarial training for adverse conditions: Robust metric localisation using appearance transfer |
| title | Adversarial training for adverse conditions: Robust metric localisation using appearance transfer |
| title_full | Adversarial training for adverse conditions: Robust metric localisation using appearance transfer |
| title_fullStr | Adversarial training for adverse conditions: Robust metric localisation using appearance transfer |
| title_full_unstemmed | Adversarial training for adverse conditions: Robust metric localisation using appearance transfer |
| title_short | Adversarial training for adverse conditions: Robust metric localisation using appearance transfer |
| title_sort | adversarial training for adverse conditions robust metric localisation using appearance transfer |
| work_keys_str_mv | AT poravh adversarialtrainingforadverseconditionsrobustmetriclocalisationusingappearancetransfer AT maddernw adversarialtrainingforadverseconditionsrobustmetriclocalisationusingappearancetransfer AT newmanp adversarialtrainingforadverseconditionsrobustmetriclocalisationusingappearancetransfer |