R-SQAIR: Relational Sequential Attend, Infer, Repeat

Traditional sequential multi-object attention models rely on a recurrent mechanism to infer object relations. We propose a relational extension (R-SQAIR) of one such attention model (SQAIR) by endowing it with a module with strong relational inductive bias that computes in parallel pairwise interactions between inferred objects. Two recently proposed relational modules are studied on tasks of unsupervised learning from videos. We demonstrate gains over sequential relational mechanisms, also in terms of combinatorial generalization.

[1]  Alexander Lerchner,et al.  COBRA: Data-Efficient Model-Based RL through Unsupervised Object Discovery and Curiosity-Driven Exploration , 2019, ArXiv.

[2]  Jürgen Schmidhuber,et al.  Recurrent World Models Facilitate Policy Evolution , 2018, NeurIPS.

[3]  Jürgen Schmidhuber,et al.  An on-line algorithm for dynamic reinforcement learning and planning in reactive environments , 1990, 1990 IJCNN International Joint Conference on Neural Networks.

[4]  Jürgen Schmidhuber,et al.  Relational Neural Expectation Maximization: Unsupervised Discovery of Objects and their Interactions , 2018, ICLR.

[5]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[6]  E. Spelke,et al.  Perception of partly occluded objects in infancy , 1983, Cognitive Psychology.

[7]  Matthew Botvinick,et al.  MONet: Unsupervised Scene Decomposition and Representation , 2019, ArXiv.

[8]  Jason Weston,et al.  End-To-End Memory Networks , 2015, NIPS.

[9]  Kristian Kersting,et al.  Faster Attend-Infer-Repeat with Tractable Probabilistic Models , 2019, ICML.

[10]  Ruslan Salakhutdinov,et al.  Importance Weighted Autoencoders , 2015, ICLR.

[11]  Razvan Pascanu,et al.  Visual Interaction Networks: Learning a Physics Simulator from Video , 2017, NIPS.

[12]  Sergio Gomez Colmenarejo,et al.  Hybrid computing using a neural network with dynamic external memory , 2016, Nature.

[13]  Klaus Greff,et al.  Multi-Object Representation Learning with Iterative Variational Inference , 2019, ICML.

[14]  Razvan Pascanu,et al.  Relational inductive biases, deep learning, and graph networks , 2018, ArXiv.

[15]  Geoffrey E. Hinton,et al.  Attend, Infer, Repeat: Fast Scene Understanding with Generative Models , 2016, NIPS.

[16]  P. Cochat,et al.  Et al , 2008, Archives de pediatrie : organe officiel de la Societe francaise de pediatrie.

[17]  Alexander Ilin,et al.  Recurrent Ladder Networks , 2017, NIPS.

[18]  J. Schmidhuber.  Reducing the Ratio Between Learning Complexity and Number of Time Varying Variables in Fully Recurrent Nets , 1993 .

[19]  Jürgen Schmidhuber,et al.  Learning to Generate Artificial Fovea Trajectories for Target Detection , 1991, Int. J. Neural Syst..

[20]  Alex Graves,et al.  Neural Turing Machines , 2014, ArXiv.

[21]  Lukasz Kaiser,et al.  Attention is All you Need , 2017, NIPS.

[22]  Klaus Greff,et al.  A Perspective on Objects and Systematic Generalization in Model-Based RL , 2019, ArXiv.

[23]  S. Carey,et al.  Infants’ Metaphysics: The Case of Numerical Identity , 1996, Cognitive Psychology.

[24]  Juan Carlos Niebles,et al.  Learning to Decompose and Disentangle Representations for Video Prediction , 2018, NeurIPS.

[25]  E. Spelke,et al.  Spatiotemporal continuity, smoothness of motion and object identity in infancy , 1995 .

[26]  Jürgen Schmidhuber,et al.  Neural Expectation Maximization , 2017, NIPS.

[27]  Yee Whye Teh,et al.  Sequential Attend, Infer, Repeat: Generative Modelling of Moving Objects , 2018, NeurIPS.

[29]  Joelle Pineau,et al.  Spatially Invariant Unsupervised Object Detection with Convolutional Neural Networks , 2019, AAAI.

[30]  Harri Valpola,et al.  Tagger: Deep Unsupervised Perceptual Grouping , 2016, NIPS.

[31]  Razvan Pascanu,et al.  Interaction Networks for Learning about Objects, Relations and Physics , 2016, NIPS.

[32]  Tapani Raiko,et al.  Semi-supervised Learning with Ladder Networks , 2015, NIPS.

[33]  Bin Li,et al.  Generative Modeling of Infinite Occluded Objects for Compositional Scene Representation , 2019, ICML.

[34]  S. Carey,et al.  The perception of causality in infancy. , 2006, Acta psychologica.

[35]  Razvan Pascanu,et al.  Relational recurrent neural networks , 2018, NeurIPS.

[36]  H. Furth.  Object permanence in five-month-old infants. , 1987, Cognition.