Optimal Transport and Deep Generative Models

59 mins 3 secs,  225.86 MB,  iPod Video  480x270,  29.97 fps,  44100 Hz,  522.22 kbits/sec
Share this media item:
Embed this media item:


About this item
Image inherited from collection
Description: Peyre, G (CNRS - Ecole Normale Superieure Paris)
Thursday 14th December 2017 - 10:00 to 11:00
 
Created: 2017-12-15 15:46
Collection: Variational methods and effective algorithms for imaging and vision
Publisher: Isaac Newton Institute
Copyright: Peyre, G
Language: eng (English)
Distribution: World     (downloadable)
Explicit content: No
Aspect Ratio: 16:9
Screencast: No
Bumper: UCS Default
Trailer: UCS Default
 
Abstract: Co-authors: Marco Cuturi (ENSAE), Aude Genevay (ENS)

In this talk, I will review some recent advances on deep generative models through the prism of Optimal Transport (OT). OT provides a way to define robust loss functions to perform high dimensional density fitting using generative models. This defines so called Minimum Kantorovitch Estimators (MKE) [1]. This approach is especially useful to recast several unsupervised deep learning methods in a unifying framework. Most notably, as shown respectively in [2,3] (and reviewed in [4]) Variational Autoencoders (VAE) and Generative Adversarial Networks (GAN) can be interpreted as (respectively primal and and dual) approximate MKE. This is a joint work with Aude Genevay and Marco Cuturi.

References: [1] Federico Bassetti, Antonella Bodini, and Eugenio Regazzini. On minimum Kantorovich distance estimators. Statistics & probability letters, 76(12):1298–1302, 2006. [2] Olivier Bousquet, Sylvain Gelly, Ilya Tolstikhin, Carl-Johann Simon-Gabriel, and Bernhard Schoelkopf. From optimal transport to generative modeling: the VEGAN cookbook. Arxiv:1705.07642, 2017. [3] Martin Arjovsky, Soumith Chintala, and Léon Bottou. Wasserstein GAN. Arxiv:1701.07875, 2017. [4] Aude Genevay, Gabriel Peyré, Marco Cuturi, GAN and VAE from an Optimal Transport Point of View, Arxiv:1706.01807, 2017

Related Links

https://arxiv.org/abs/1706.01807 - GAN and VAE from an Optimal Transport Point of View
Available Formats
Format Quality Bitrate Size
MPEG-4 Video 640x360    1.93 Mbits/sec 857.64 MB View Download
WebM 640x360    771.33 kbits/sec 333.69 MB View Download
iPod Video * 480x270    522.22 kbits/sec 225.86 MB View Download
MP3 44100 Hz 249.73 kbits/sec 108.13 MB Listen Download
Auto (Allows browser to choose a format it supports)