Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2023
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2310.10410 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913224188755968 |
|---|---|
| author | Traub, Manuel Becker, Frederic Sauter, Adrian Otte, Sebastian Butz, Martin V. |
| author_facet | Traub, Manuel Becker, Frederic Sauter, Adrian Otte, Sebastian Butz, Martin V. |
| contents | Current slot-oriented approaches for compositional scene segmentation from images and videos rely on provided background information or slot assignments. We present a segmented location and identity tracking system, Loci-Segmented (Loci-s), which does not require either of this information. It learns to dynamically segment scenes into interpretable background and slot-based object encodings, separating rgb, mask, location, and depth information for each. The results reveal largely superior video decomposition performance in the MOVi datasets and in another established dataset collection targeting scene segmentation. The system's well-interpretable, compositional latent encodings may serve as a foundation model for downstream tasks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2310_10410 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Loci-Segmented: Improving Scene Segmentation Learning Traub, Manuel Becker, Frederic Sauter, Adrian Otte, Sebastian Butz, Martin V. Computer Vision and Pattern Recognition Current slot-oriented approaches for compositional scene segmentation from images and videos rely on provided background information or slot assignments. We present a segmented location and identity tracking system, Loci-Segmented (Loci-s), which does not require either of this information. It learns to dynamically segment scenes into interpretable background and slot-based object encodings, separating rgb, mask, location, and depth information for each. The results reveal largely superior video decomposition performance in the MOVi datasets and in another established dataset collection targeting scene segmentation. The system's well-interpretable, compositional latent encodings may serve as a foundation model for downstream tasks. |
| title | Loci-Segmented: Improving Scene Segmentation Learning |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2310.10410 |