Understanding Visual Concepts with Continuation Learning
Explore this paper's citation graph
Summary
A neural network architecture and a learning algorithm are introduced to produce factorized symbolic representations that demonstrate the efficacy of this approach on datasets of faces undergoing 3D transformations and Atari 2600 games.
- Type
- preprint
- Published
- 2016-02-22
- Cited by
- 55
- References
- 17
- Access
- Open access
- OpenAlex
- https://openalex.org/W2281112906
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:17914446
Keywords
Continuation, Frame (networking), Representation (politics), Computer science, Set (abstract data type)
References
- Deep Convolutional Inverse Graphics Network
- Stochastic Backpropagation and Approximate Inference in Deep Generative Models
- Auto-Encoding Variational Bayes
- Learning Deep Architectures for AI
- A 3D Face Model for Pose and Illumination Invariant Face Recognition
- Convolutional deep belief networks for scalable unsupervised learning of hierarchical representations
- Unsupervised Learning of Invariant Feature Hierarchies with Applications to Object Recognition
- Human-level control through deep reinforcement learning
- Human-level concept learning through probabilistic program induction
- Disentangled Representations in Neural Models
- Robot vision
- Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
- Reducing the Dimensionality of Data with Neural Networks
- Neural Turing Machines
- Transforming Auto-Encoders
- Stacked Convolutional Auto-Encoders for Hierarchical Feature Extraction
- This PDF file includes: Materials and Methods
Cited by
- Hierarchical Deep Reinforcement Learning: Integrating Temporal Abstraction and Intrinsic Motivation
- Deep Successor Reinforcement Learning
- InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets
- Early Visual Concept Learning with Unsupervised Deep Learning
- ExtremeWeather: A large-scale climate dataset for semi-supervised detection, localization, and understanding of extreme weather events
- Toward an Integration of Deep Learning and Neuroscience
- Time-Contrastive Networks: Self-Supervised Learning from Multi-view Observation
- beta-VAE: Learning Basic Visual Concepts with a Constrained Variational Framework
- Variational Inference of Disentangled Latent Concepts from Unlabeled Observations
- Time-Contrastive Networks: Self-Supervised Learning from Video
- A Framework for the Quantitative Evaluation of Disentangled Representations
- Understanding disentangling in β-VAE
- Conditional Image Generation for Learning the Structure of Visual Objects
- A Comprehensive survey on deep future frame video prediction
- Unsupervised Learning of Object Landmarks through Conditional Image Generation
- Image Classification Using Deep Autoencoders
- Towards a Definition of Disentangled Representations
- Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations
- Disentangling Video with Independent Prediction
- Relevance Factor VAE: Learning and Identifying Disentangled Factors
Related papers
- Deep Convolutional Inverse Graphics Network
- beta-VAE: Learning Basic Visual Concepts with a Constrained Variational Framework
- Learning to Disentangle Factors of Variation with Manifold Interaction
- Representation Learning: A Review and New Perspectives
- Weakly-supervised Disentangling with Recurrent Transformations for 3D View Synthesis
- Auto-Encoding Variational Bayes