Timezone: »
Convolutional neural networks provide visual features that perform remarkably well in many computer vision applications. However, training these networks requires significant amounts of supervision; this paper introduces a generic framework to train such networks, end-to-end, with no supervision. We propose to fix a set of target representations, called Noise As Targets (NAT), and to constrain the deep features to align to them. This domain agnostic approach avoids the standard unsupervised learning issues of trivial solutions and collapsing of the features. Thanks to a stochastic batch reassignment strategy and a separable square loss function, it scales to millions of images. The proposed approach produces representations that perform on par with the state-of-the-arts among unsupervised methods on ImageNet and Pascal VOC.
Summary/Notes]
Author Information
Piotr Bojanowski (Facebook)
Armand Joulin (Facebook)
Related Events (a corresponding poster, oral, or spotlight)
-
2017 Poster: Unsupervised Learning by Predicting Noise »
Mon. Aug 7th 08:30 AM -- 12:00 PM Room Gallery #76
More from the Same Authors
-
2018 Poster: Optimizing the Latent Space of Generative Networks »
Piotr Bojanowski · Armand Joulin · David Lopez-Paz · Arthur Szlam -
2018 Oral: Optimizing the Latent Space of Generative Networks »
Piotr Bojanowski · Armand Joulin · David Lopez-Paz · Arthur Szlam -
2017 Poster: Efficient softmax approximation for GPUs »
Edouard Grave · Armand Joulin · Moustapha Cisse · David Grangier · Herve Jegou -
2017 Poster: Parseval Networks: Improving Robustness to Adversarial Examples »
Moustapha Cisse · Piotr Bojanowski · Edouard Grave · Yann Dauphin · Nicolas Usunier -
2017 Talk: Efficient softmax approximation for GPUs »
Edouard Grave · Armand Joulin · Moustapha Cisse · David Grangier · Herve Jegou -
2017 Talk: Parseval Networks: Improving Robustness to Adversarial Examples »
Moustapha Cisse · Piotr Bojanowski · Edouard Grave · Yann Dauphin · Nicolas Usunier