Conference item : Poster
Reconstruction bottlenecks in object-centric generative models
- Abstract:
- A range of methods with suitable inductive biases exist to learn interpretable object-centric representations of images without supervision. However, these are largely restricted to visually simple images; robust object discovery in real-world sensory datasets remains elusive. To increase the understanding of such inductive biases, we empirically investigate the role of “reconstruction bottlenecks” for scene decomposition in GENESIS, a recent VAE-based model. We show such bottlenecks determine reconstruction and segmentation quality and critically influence model behaviour.
- Publication status:
- Published
- Peer review status:
- Reviewed (other)
Actions
Access Document
- Files:
-
-
(Preview, Version of record, pdf, 577.7KB, Terms of use)
-
- Publication website:
- https://oolworkshop.github.io/program/ool_5.html
Authors
- Publisher:
- International Conference on Machine Learning
- Publication date:
- 2020-07-17
- Acceptance date:
- 2020-06-01
- Event title:
- Workshop on Object-Oriented Learning at ICML 2020
- Event series:
- International Conference on Machine Learning
- Event location:
- Online
- Event website:
- https://icml.cc/Conferences/2020
- Event start date:
- 2020-07-12
- Event end date:
- 2020-07-18
- Language:
-
English
- Keywords:
- Subtype:
-
Poster
- Pubs id:
-
1138269
- Local pid:
-
pubs:1138269
- Deposit date:
-
2020-10-19
- ARK identifier:
Terms of use
- Copyright holder:
- Engelcke et al.
- Copyright date:
- 2020
- Rights statement:
- Copyright 2020 by the author(s).
If you are the owner of this record, you can report an update to it here: Report update to this record