Hadjivelichkov, Denis;
(2025)
Semantic Correspondence in Robot Perception.
Doctoral thesis (Ph.D), UCL (University College London).
Preview |
Text
_Dissertation__Dennis___Correspondences-4.pdf - Published Version Download (62MB) | Preview |
Abstract
The ability of robots to discern and leverage semantic similarities within their environment is pivotal for achieving robust perception, interaction, and autonomy. However, this task presents significant challenges due to the inherent complexity and variability of unstructured real environments, where objects exhibit diverse appearances. Traditional supervised learning approaches are not well-suited for semantic point and part correspondence, as ground truth labels are ambiguous and difficult to acquire at scale. This dissertation tackles the significant challenges associated with inferring semantic correspondences among objects within diverse scenes. Specifically, we introduce a novel technique for generating class-aware dense pixel features, aiming to enhance the granularity and robustness of object representation. Further, we introduce the first method for a one-shot visual semantic search of object parts. By leveraging these established part correspondences, we demonstrate how just one or a few examples of object interactions can be used to directly predict how a completely new pair of objects might be used together. With real robot manipulators, we demonstrate the usefulness of these methods. This dissertation tackles the significant challenges associated with inferring semantic correspondences among objects within diverse scenes. Specifically, we introduce a novel technique for generating class-aware dense pixel features, aiming to enhance the granularity and robustness of object representation. Further, we introduce the first method for a one-shot visual semantic search of object parts. By leveraging these established part correspondences, we demonstrate how just one or a few examples of object interactions can be used to directly predict how a completely new pair of objects might be used together. With real robot manipulators, we demonstrate the usefulness of these methods.
| Type: | Thesis (Doctoral) |
|---|---|
| Qualification: | Ph.D |
| Title: | Semantic Correspondence in Robot Perception |
| Open access status: | An open access version is available from UCL Discovery |
| Language: | English |
| Additional information: | Copyright © The Author 2025. Original content in this thesis is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) Licence (https://creativecommons.org/licenses/by-nc/4.0/). Any third-party copyright material present remains the property of its respective owner(s) and is licensed under its existing terms. Access may initially be restricted at the author’s request. |
| UCL classification: | UCL > Provost and Vice Provost Offices > UCL BEAMS UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science > Dept of Computer Science UCL |
| URI: | https://discovery.ucl.ac.uk/id/eprint/10211752 |
Archive Staff Only
 |
View Item |
