mlsanityData to 7 Oct 2026

SPACE: Unsupervised Object-Oriented Scene Representation via Spatial Attention and Decomposition

arXiv 2001.02407 ↗First posted 8 Jan 2020cs.LG cs.CV eess.IVcited by 190

Not on the map. Why?

Abstract (arXiv)

The ability to decompose complex multi-object scenes into meaningful abstractions like objects is fundamental to achieve higher-level cognition. Previous approaches for unsupervised object-oriented scene representation learning are either based on spatial-attention or scene-mixture approaches and limited in scalability which is a main obstacle towards modeling real-world scenes. In this paper, we propose a generative latent variable model, called SPACE, that provides a unified probabilistic modeling framework that combines the best of spatial-attention and scene-mixture approaches. SPACE can explicitly provide factorized object representations for foreground objects while also decomposing background segments of complex morphology. Previous models are good at either of these, but not both. SPACE also resolves the scalability problems of previous methods by incorporating parallel spatial-attention and thus is applicable to scenes with a large number of objects without performance degradations. We show through experiments on Atari and 3D-Rooms that SPACE achieves the above properties consistently in comparison to SPAIR, IODINE, and GENESIS. Results of our experiments can be found on our project website: https://sites.google.com/view/space-project-page

Key facts. Cited by 190 papers in the map (data through 7 Oct 2026). First posted on arXiv 8 Jan 2020.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 36; 2022: 39; 2023: 44; 2024: 37; 2025: 14; 2026: 20.

190 papers.

Sort:

By area:

  1. 2021Conditional Object-Centric Learning from Videocited by 247
  2. 2022Bridging the Gap to Real-World Object-Centric Learningcited by 219
  3. 2021Towards Causal Representation Learningcited by 207
  4. 2022Deep Spectral Methods: A Surprisingly Strong Baseline for Unsupervised Semantic Segmentation and Localizationcited by 150
  5. 2021Illiterate DALL-E Learns to Composecited by 147
  6. 2022SlotFormer: Unsupervised Visual Dynamics Simulation with Object-Centric Modelscited by 146
  7. 2022Simple Unsupervised Object-Centric Learning for Complex and Naturalistic Videoscited by 139
  8. 2021GENESIS-V2: Inferring Unordered Object Representations without Iterative Refinementcited by 109
  9. 2021Unsupervised Discovery of Object Radiance Fieldscited by 100
  10. 2021ClevrTex: A Texture-Rich Benchmark for Unsupervised Multi-Object Segmentationcited by 93
  11. 2022Object Scene Representation Transformercited by 90
  12. 2023Object-Centric Slot Diffusioncited by 90
38 more
  1. 2021Decomposing 3D Scenes into Objects via Unsupervised Volume Segmentationcited by 88
  2. 2023SlotDiffusion: Object-Centric Generative Modeling with Diffusion Modelscited by 85
  3. 2021Generalization and Robustness Implications in Object-Centric Learningcited by 81
  4. 2023Object-Centric Learning for Real-World Videos by Predicting Temporal Feature Similaritiescited by 78
  5. 2021SIMONe: View-Invariant, Temporally-Abstracted Object Representations via Unsupervised Video Decompositioncited by 69
  6. 2022Improving Object-centric Learning with Query Optimizationcited by 60
  7. 2021Unsupervised Part Discovery from Contrastive Reconstructioncited by 57
  8. 2023Additive Decoders for Latent Variables Identification and Cartesian-Product Extrapolationcited by 56
  9. 2023Invariant Slot Attention: Object Discovery with Slot-Centric Reference Framescited by 51
  10. 2021Unsupervised Layered Image Decomposition into Object Prototypescited by 50
  11. 2023An Investigation into Pre-Training Object-Centric Representations for Reinforcement Learningcited by 49
  12. 2023Self-supervised Object-Centric Learning for Videoscited by 49
  13. 2023SPOT: Self-Training with Patch-Order Permutation for Object-Centric Learning with Autoregressive Transformerscited by 49
  14. 2023Read and Reap the Rewards: Learning to Play Atari with the Help of Instruction Manualscited by 45
  15. 2023Provably Learning Object-Centric Representationscited by 45
  16. 2023Interpretable and Explainable Logical Policies via Neurally Guided Symbolic Abstractioncited by 45
  17. 2022Discovering Objects that Can Movecited by 44
  18. 2022Guess What Moves: Unsupervised Video and Image Segmentation by Anticipating Motioncited by 42
  19. 2022Compositional Generalization in Unsupervised Compositional Representation Learning: A Study on Disentanglement and Emergent Languagecited by 42
  20. 2023Provable Compositional Generalization for Object-Centric Learningcited by 41
  21. 2022Complex-Valued Autoencoders for Object Discoverycited by 40
  22. 2021Learning Object-Centric Representations of Multi-Object Scenes from Multiple Viewscited by 37
  23. 2022Neural Systematic Bindercited by 37
  24. 2021Efficient Iterative Amortized Inference for Learning Symmetric and Disentangled Multi-Object Representationscited by 36
  25. 2021MarioNette: Self-Supervised Sprite Learningcited by 35
  26. 2022VAEL: Bridging Variational Autoencoders and Probabilistic Logic Programmingcited by 31
  27. 2023OCAtari: Object-Centric Atari 2600 Reinforcement Learning Environmentscited by 30
  28. 2021Unsupervised Object Learning via Common Fatecited by 29
  29. 2023Rotating Features for Object Discoverycited by 29
  30. 2022Unsupervised Discovery and Composition of Object Light Fieldscited by 28
  31. 2022Compositional Scene Representation Learning via Reconstruction: A Surveycited by 27
  32. 2021Unsupervised Foreground Extraction via Deep Region Competitioncited by 26
  33. 2022Reasoning with Scene Graphs for Robot Planning under Partial Observabilitycited by 25
  34. 2023Object Discovery from Motion-Guided Tokenscited by 25
  35. 2021Generative Video Transformer: Can Objects be the Words?cited by 24
  36. 2021Compositional Transformers for Scene Generationcited by 24
  37. 2021Interactive Disentanglement: Learning Concepts by Interacting with their Prototype Representationscited by 24
  38. 2021Adaptive Rational Activations to Boost Deep Reinforcement Learningcited by 23

50 most cited of 190 All 190 (JSON)