Open AccessPosted Content
Visual Representation Learning Does Not Generalize Strongly Within the Same Domain
Lukas Schott,Julius von Kügelgen,Frederik Träuble,Peter V. Gehler,Chris Russell,Matthias Bethge,Bernhard Schölkopf,Francesco Locatello,Wieland Brendel +8 more
Reads0
Chats0
TLDR
This paper test whether 17 unsupervised, weakly supervised, and fully supervised representation learning approaches correctly infer the generative factors of variation in simple datasets and observe that all of them struggle to learn the underlying mechanism regardless of supervision signal and architectural bias.Abstract:
An important component for generalization in machine learning is to uncover underlying latent factors of variation as well as the mechanism through which each factor acts in the world. In this paper, we test whether 17 unsupervised, weakly supervised, and fully supervised representation learning approaches correctly infer the generative factors of variation in simple datasets (dSprites, Shapes3D, MPI3D). In contrast to prior robustness work that introduces novel factors of variation during test time, such as blur or other (un)structured noise, we here recompose, interpolate, or extrapolate only existing factors of variation from the training data set (e.g., small and medium-sized objects during training and large objects during testing). Models that learn the correct mechanism should be able to generalize to this benchmark. In total, we train and test 2000+ models and observe that all of them struggle to learn the underlying mechanism regardless of supervision signal and architectural bias. Moreover, the generalization capabilities of all tested models drop significantly as we move from artificial datasets towards more realistic real-world datasets. Despite their inability to identify the correct mechanism, the models are quite modular as their ability to infer other in-distribution factors remains fairly stable, providing only a single factor is out-of-distribution. These results point to an important yet understudied problem of learning mechanistic models of observations that can facilitate generalization.read more
Citations
More filters
Journal ArticleDOI
How we learn: why brains learn better than any machine … for now: by Stanislas Dehaene, New York, Viking, 2020, £20.15 (Hardback), £12.23 (Paperback), ISBN: 9780525559887
TL;DR: Dehaene et al. as discussed by the authors used a quote from MIT President L. Rafael Reif: "If we don't know how to learn, we can't learn how to adapt."
Proceedings ArticleDOI
Assaying Out-Of-Distribution Generalization in Transfer Learning
Florian Wenzel,Andrea Dittadi,Peter V. Gehler,Carl-Johann Simon-Gabriel,Max Horn,Dominik Zietlow,David Kernert,Chris Russell,Thomas Brox,Bernt Schiele,Bernhard Schölkopf,Francesco Locatello +11 more
TL;DR: A view of previous work is taken, highlighting message discrepancies that are addressed empirically, and providing recommendations on how to measure the robustness of a model and how to improve it, to gain a broader insight in the sometimes contradicting statements on OOD robustness in previous research.
Journal ArticleDOI
Controlled Generation of Unseen Faults for Partial and OpenSet&Partial Domain Adaptation
TL;DR: A new framework for Partial and Open-Partial domain adaptation based on generating distinct fault signatures with a Wasserstein GAN is proposed, which is especially suited in extreme domain adaption settings that are particularly relevant in the context of complex and safety-critical systems.
Proceedings ArticleDOI
Disentangling with Biological Constraints: A Theory of Functional Cell Types
TL;DR: This work mathematically proves that simple biological constraints on neurons, namely nonnegativity and energy in both activity and weights, promote such sought after disentangled representations by enforcing neurons to become selective for single factors of task variation.
Journal ArticleDOI
Identifiability of deep generative models under mixture priors without auxiliary information
TL;DR: This work proves identifiability of a broad class of deep latent variable models that (a) have universal approximation capabilities and (b) are the decoders of variational autoencoders that are commonly used in practice.
References
More filters
Proceedings ArticleDOI
Deep Residual Learning for Image Recognition
TL;DR: In this article, the authors proposed a residual learning framework to ease the training of networks that are substantially deeper than those used previously, which won the 1st place on the ILSVRC 2015 classification task.
Proceedings Article
ImageNet Classification with Deep Convolutional Neural Networks
TL;DR: The state-of-the-art performance of CNNs was achieved by Deep Convolutional Neural Networks (DCNNs) as discussed by the authors, which consists of five convolutional layers, some of which are followed by max-pooling layers, and three fully-connected layers with a final 1000-way softmax.
Proceedings ArticleDOI
ImageNet: A large-scale hierarchical image database
TL;DR: A new database called “ImageNet” is introduced, a large-scale ontology of images built upon the backbone of the WordNet structure, much larger in scale and diversity and much more accurate than the current image datasets.
Journal ArticleDOI
Deep learning
TL;DR: Deep learning is making major advances in solving problems that have resisted the best attempts of the artificial intelligence community for many years, and will have many more successes in the near future because it requires very little engineering by hand and can easily take advantage of increases in the amount of available computation and data.
Book
The Nature of Statistical Learning Theory
TL;DR: Setting of the learning problem consistency of learning processes bounds on the rate of convergence ofLearning processes controlling the generalization ability of learning process constructing learning algorithms what is important in learning theory?
Related Papers (5)
Global-Local Bidirectional Reasoning for Unsupervised Representation Learning of 3D Point Clouds.
Yongming Rao,Jiwen Lu,Jie Zhou +2 more