mlsanityData to 14 Oct 2026

Identifiable Object Representations under Spatial Ambiguities

arXiv 2506.07806 ↗First posted 9 Jun 2025cs.LG cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Modular object-centric representations are essential for *human-like reasoning* but are challenging to obtain under spatial ambiguities, *e.g. due to occlusions and view ambiguities*. However, addressing challenges presents both theoretical and practical difficulties. We introduce a novel multi-view probabilistic approach that aggregates view-specific slots to capture *invariant content* information while simultaneously learning disentangled global *viewpoint-level* information. Unlike prior single-view methods, our approach resolves spatial ambiguities, provides theoretical guarantees for identifiability, and requires *no viewpoint annotations*. Extensive experiments on standard benchmarks and novel complex datasets validate our method's robustness and scalability.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 54 of its 76 references that are in the map. First posted on arXiv 9 Jun 2025.

Builds on

54 of 76 references.

References · 54

  1. 2017Attention Is All You Needcited by 60,361
  2. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,968
  3. 2015Spatial Transformer Networkscited by 1,429
  4. 2018Relational inductive biases, deep learning, and graph networkscited by 1,360
  5. 2018Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representationscited by 941
  6. 2020Object-Centric Learning with Slot Attentioncited by 876
  7. 2018Disentangling by Factorisingcited by 616
  8. 2022Google Scanned Objects: A High-Quality Dataset of 3D Scanned Household Itemscited by 612
  9. 2016Learning a Probabilistic Latent Space of Object Shapes via 3D Generative-Adversarial Modelingcited by 558
  10. 2019Variational Autoencoders and Nonlinear ICA: A Unifying Frameworkcited by 503
  11. 2019MONet: Unsupervised Scene Decomposition and Representationcited by 392
  12. 2022Kubric: A scalable dataset generatorcited by 381
38 more
  1. 2019Multi-Object Representation Learning with Iterative Variational Inferencecited by 360
  2. 2021Self-Supervised Learning with Data Augmentations Provably Isolates Content from Stylecited by 312
  3. 2018Nonlinear ICA Using Auxiliary Variables and Generalized Contrastive Learningcited by 282
  4. 2021Conditional Object-Centric Learning from Videocited by 248
  5. 2020Weakly-Supervised Disentanglement Without Compromisescited by 239
  6. 2020On the Binding Problem in Artificial Neural Networkscited by 233
  7. 2019GENESIS: Generative Scene Inference and Sampling with Object-Centric Latent Representationscited by 227
  8. 2021Contrastive Learning Inverts the Data Generating Processcited by 221
  9. 2016Deep Unsupervised Clustering with Gaussian Mixture Variational Autoencoderscited by 220
  10. 2022Bridging the Gap to Real-World Object-Centric Learningcited by 220
  11. 2020SPACE: Unsupervised Object-Oriented Scene Representation via Spatial Attention and Decompositioncited by 190
  12. 2021Illiterate DALL-E Learns to Composecited by 147
  13. 2022Interventional Causal Representation Learningcited by 142
  14. 2022Weakly supervised causal representation learningcited by 140
  15. 2018Sequential Attend, Infer, Repeat: Generative Modelling of Moving Objectscited by 128
  16. 2021How to represent part-whole hierarchies in a neural networkcited by 123
  17. 2018Learning to Decompose and Disentangle Representations for Video Predictioncited by 121
  18. 2021GENESIS-V2: Inferring Unordered Object Representations without Iterative Refinementcited by 109
  19. 2020ICE-BeeM: Identifiable Conditional Energy-Based Deep Models Based on Nonlinear ICAcited by 102
  20. 2018Relational Neural Expectation Maximization: Unsupervised Discovery of Objects and their Interactionscited by 100
  21. 2018Iterative Amortized Inferencecited by 94
  22. 2019The Incomplete Rosetta Stone Problem: Identifiability Results for Multi-View Nonlinear ICAcited by 86
  23. 2022Identifiability of deep generative models without auxiliary informationcited by 70
  24. 2023Identifiability Results for Multimodal Contrastive Learningcited by 61
  25. 2023Additive Decoders for Latent Variables Identification and Cartesian-Product Extrapolationcited by 58
  26. 2022Function Classes for Identifiable Nonlinear Independent Component Analysiscited by 53
  27. 2022Object Representations as Fixed Points: Training Iterative Refinement Algorithms with Implicit Differentiationcited by 46
  28. 2023Provably Learning Object-Centric Representationscited by 45
  29. 2019Exploiting Spatial Invariance for Scalable Unsupervised Object Trackingcited by 40
  30. 2020ROOTS: Object-Centric Representation and Rendering of 3D Scenescited by 40
  31. 2022From Statistical to Causal Learningcited by 40
  32. 2021Learning Object-Centric Representations of Multi-Object Scenes from Multiple Viewscited by 37
  33. 2023Rotating Features for Object Discoverycited by 29
  34. 2021I Don't Need u: Identifiable Non-Linear ICA Without Side Informationcited by 21
  35. 2023Object-centric architectures enable efficient causal representation learningcited by 21
  36. 2023Slot-VAE: Object-Centric Scene Generation with Slot Attentioncited by 20
  37. 2023Grounded Object Centric Learningcited by 17
  38. 2024Unsupervised Discovery of Object-Centric Neural Fieldscited by 12

50 most cited of 54 All 54 (JSON)

Built on by

Not cited yet.