Multimodal Variational Autoencoders for Semi-Supervised Learning: In Defense of Product-of-Experts
Authors:
Svetlana Kutuzova,
Oswin Krause,
Douglas McCloskey,
Mads Nielsen,
Christian Igel
Abstract:
Multimodal generative models should be able to learn a meaningful latent representation that enables a coherent joint generation of all modalities (e.g., images and text). Many applications also require the ability to accurately sample modalities conditioned on observations of a subset of the modalities. Often not all modalities may be observed for all training data points, so semi-supervised lear…
▽ More
Multimodal generative models should be able to learn a meaningful latent representation that enables a coherent joint generation of all modalities (e.g., images and text). Many applications also require the ability to accurately sample modalities conditioned on observations of a subset of the modalities. Often not all modalities may be observed for all training data points, so semi-supervised learning should be possible. In this study, we propose a novel product-of-experts (PoE) based variational autoencoder that have these desired properties. We benchmark it against a mixture-of-experts (MoE) approach and an approach of combining the modalities with an additional encoder network. An empirical evaluation shows that the PoE based models can outperform the contrasted models. Our experiments support the intuition that PoE models are more suited for a conjunctive combination of modalities.
△ Less
Submitted 30 July, 2021; v1 submitted 18 January, 2021;
originally announced January 2021.
A Many Antenna High Rate Wireless System
Authors:
Philip Gossett,
Jeremy Thorpe,
Bob Nuckolls,
Brett Coon,
Dan McCloskey,
David Chang,
Greg Steuck,
Paul Rodman,
Sasha Levitskiy,
Yuan Yuan
Abstract:
We describe a TDD MIMO wireless system designed to operate at high bandwidth and low SNR. Signals are transmitted as a direct sequence. In the uplink (Multiple Access Channel), signal detection is done by a space-time whitening filter followed by a matched filter. In the downlink (Broadcast Channel), precoding is done by the transpose of these filters.
We further describe an implementation of th…
▽ More
We describe a TDD MIMO wireless system designed to operate at high bandwidth and low SNR. Signals are transmitted as a direct sequence. In the uplink (Multiple Access Channel), signal detection is done by a space-time whitening filter followed by a matched filter. In the downlink (Broadcast Channel), precoding is done by the transpose of these filters.
We further describe an implementation of this system that uses an array of 32 antennas to communicate with 32 single-antenna clients simultaneously on the same frequencies between 512-608 and 614-698 MHz. At close range, all 32 links achieve the full PHY data rate, both uplink and downlink, with less than 1\% Block Error Rate on each link. The total system rate is 3.8 Gb/s. The system spectral efficiency is 21.7 b/s/Hz for both uplink and downlink.
We close with some projections to the not-to-distant future.
△ Less
Submitted 5 May, 2016;
originally announced May 2016.
