Unsupervised Feature Learning for Manipulation with Contrastive Domain Randomization

1 code implementation • ICLR Workshop SSL-RL 2021 • Carmel Rabinovitz, Niko Grupen, Aviv Tamar

In this work, however, we show that a naive application of DR to unsupervised learning based on contrastive estimation does not promote invariance, as the loss function maximizes mutual information between the features and both the relevant and irrelevant visual properties.

PERL: Pivot-based Domain Adaptation for Pre-trained Deep Contextualized Embedding Models

1 code implementation • 16 Jun 2020 • Eyal Ben-David, Carmel Rabinovitz, Roi Reichart

Pivot-based neural representation models have lead to significant progress in domain adaptation for NLP.

Domain Adaptation Representation Learning +2

High quality, lightweight and adaptable TTS using LPCNet

no code implementations • 2 May 2019 • Zvi Kons, Slava Shechtman, Alex Sorin, Carmel Rabinovitz, Ron Hoory

We first demonstrate the ability of the system to produce high quality speech when trained on large, high quality datasets.

Audio and Speech Processing Sound