Identifiable Object-Centric Representation Learning via Probabilistic Slot Attention
Abstract
Learning modular object-centric representations is said to be crucial for systematic generalization. Existing methods show promising object-binding capabilities empirically, but theoretical identifiability guarantees remain relatively underdeveloped. Understanding when object-centric representations can theoretically be identified is important for scaling slot-based methods to high-dimensional images with correctness guarantees. To that end, we propose a probabilistic slot-attention algorithm that imposes an aggregate mixture prior over object-centric slot representations, thereby providing slot identifiability guarantees without supervision, up to an equivalence relation. We provide empirical verification of our theoretical identifiability result using both simple 2-dimensional data and high-resolution imaging datasets.
Cite
Text
Kori et al. "Identifiable Object-Centric Representation Learning via Probabilistic Slot Attention." Neural Information Processing Systems, 2024. doi:10.52202/079017-2960Markdown
[Kori et al. "Identifiable Object-Centric Representation Learning via Probabilistic Slot Attention." Neural Information Processing Systems, 2024.](https://mlanthology.org/neurips/2024/kori2024neurips-identifiable/) doi:10.52202/079017-2960BibTeX
@inproceedings{kori2024neurips-identifiable,
title = {{Identifiable Object-Centric Representation Learning via Probabilistic Slot Attention}},
author = {Kori, Avinash and Locatello, Francesco and Santhirasekaram, Ainkaran and Toni, Francesca and Glocker, Ben and Ribeiro, Fabio De Sousa},
booktitle = {Neural Information Processing Systems},
year = {2024},
doi = {10.52202/079017-2960},
url = {https://mlanthology.org/neurips/2024/kori2024neurips-identifiable/}
}