Slot-VAE: Object-Centric Scene Generation with Slot Attention

Yanbo Wang, Letao Liu, Justin Dauwels
Proceedings of the 40th International Conference on Machine Learning, PMLR 202:36020-36035, 2023.

Abstract

Slot attention has shown remarkable object-centric representation learning performance in computer vision tasks without requiring any supervision. Despite its object-centric binding ability brought by compositional modelling, as a deterministic module, slot attention lacks the ability to generate novel scenes. In this paper, we propose the Slot-VAE, a generative model that integrates slot attention with the hierarchical VAE framework for object-centric structured scene generation. For each image, the model simultaneously infers a global scene representation to capture high-level scene structure and object-centric slot representations to embed individual object components. During generation, slot representations are generated from the global scene representation to ensure coherent scene structures. Our extensive evaluation of the scene generation ability indicates that Slot-VAE outperforms slot representation-based generative baselines in terms of sample quality and scene structure accuracy.

Cite this Paper

BibTeX
@InProceedings{pmlr-v202-wang23r, title = {Slot-{VAE}: Object-Centric Scene Generation with Slot Attention}, author = {Wang, Yanbo and Liu, Letao and Dauwels, Justin}, booktitle = {Proceedings of the 40th International Conference on Machine Learning}, pages = {36020--36035}, year = {2023}, editor = {Krause, Andreas and Brunskill, Emma and Cho, Kyunghyun and Engelhardt, Barbara and Sabato, Sivan and Scarlett, Jonathan}, volume = {202}, series = {Proceedings of Machine Learning Research}, month = {23--29 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v202/wang23r/wang23r.pdf}, url = {https://proceedings.mlr.press/v202/wang23r.html}, abstract = {Slot attention has shown remarkable object-centric representation learning performance in computer vision tasks without requiring any supervision. Despite its object-centric binding ability brought by compositional modelling, as a deterministic module, slot attention lacks the ability to generate novel scenes. In this paper, we propose the Slot-VAE, a generative model that integrates slot attention with the hierarchical VAE framework for object-centric structured scene generation. For each image, the model simultaneously infers a global scene representation to capture high-level scene structure and object-centric slot representations to embed individual object components. During generation, slot representations are generated from the global scene representation to ensure coherent scene structures. Our extensive evaluation of the scene generation ability indicates that Slot-VAE outperforms slot representation-based generative baselines in terms of sample quality and scene structure accuracy.} }
Endnote
%0 Conference Paper %T Slot-VAE: Object-Centric Scene Generation with Slot Attention %A Yanbo Wang %A Letao Liu %A Justin Dauwels %B Proceedings of the 40th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2023 %E Andreas Krause %E Emma Brunskill %E Kyunghyun Cho %E Barbara Engelhardt %E Sivan Sabato %E Jonathan Scarlett %F pmlr-v202-wang23r %I PMLR %P 36020--36035 %U https://proceedings.mlr.press/v202/wang23r.html %V 202 %X Slot attention has shown remarkable object-centric representation learning performance in computer vision tasks without requiring any supervision. Despite its object-centric binding ability brought by compositional modelling, as a deterministic module, slot attention lacks the ability to generate novel scenes. In this paper, we propose the Slot-VAE, a generative model that integrates slot attention with the hierarchical VAE framework for object-centric structured scene generation. For each image, the model simultaneously infers a global scene representation to capture high-level scene structure and object-centric slot representations to embed individual object components. During generation, slot representations are generated from the global scene representation to ensure coherent scene structures. Our extensive evaluation of the scene generation ability indicates that Slot-VAE outperforms slot representation-based generative baselines in terms of sample quality and scene structure accuracy.
APA
Wang, Y., Liu, L. & Dauwels, J.. (2023). Slot-VAE: Object-Centric Scene Generation with Slot Attention. Proceedings of the 40th International Conference on Machine Learning, in Proceedings of Machine Learning Research 202:36020-36035 Available from https://proceedings.mlr.press/v202/wang23r.html.

Related Material