University of Twente Student Theses
Visual Place Recognition: Building an Evaluation Framework for Model Robustness
Gosa, V.I. (2024) Visual Place Recognition: Building an Evaluation Framework for Model Robustness.
|
PDF
1MB |
| Abstract: | Visual Place Recognition (VPR) is the process of identifying and retrieving images captured at the same location as a given query image. The introduction of VPR pipelines in applications such as autonomous driving and mobile robot localization makes it crucial that models perform image retrieval tasks consistently under challenging conditions such as image blur, lossy compression, or even domain changes such as weather and time of day. Our standardized benchmark compares multiple state-of-the-art VPR pipelines on synthetically generated test datasets to mitigate the effects of uncontrollable variables caused by the image capturing process. Results show how vision transformer backbones are consistently more robust to domain changes and image corruptions compared to traditional convolutional neural network backbones. |
| Item Type: | Essay (Bachelor) |
| Faculty: | EEMCS: Electrical Engineering, Mathematics and Computer Science |
| Subject: | 54 computer science |
| Programme: | Computer Science BSc (56964) |
| Link to this item: | https://purl.utwente.nl/essays/100860 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page