Jump to Content
Omry Tuval

Omry Tuval

Omry Tuval is currently at Google Research, working on machine learning from audio and human voice. Prior to that, Omry was an engineering tech lead for Google's business intelligence and data analytics systems and tool. In previous lives, Omry was a project manager of critical data security projects, a software security researcher and, for a short while, had a great time being a Sales Engineer for a startup in the water distribution sector.
Authored Publications
Google Publications
Other Publications
Sort By
  • Title
  • Title, descending
  • Year
  • Year, descending
    Preview abstract The ultimate goal of transfer learning is to enable learning with a small amount of data, by using a strong embedding. While significant progress has been made in the visual and language domains, the speech domain does not have such a universal method. This paper presents a new representation of speech signals based on an unsupervised triplet-loss objective, which outperforms both existing state of the art and other representations on a number of transfer learning tasks in the non-semantic speech domain. The embedding is learned on a publicly available dataset, and it is tested on a variety of low-resource downstream tasks, including personalization tasks and medical domain. The model will be publicly released. View details
    Preview abstract Automatic speech recognition (ASR) systems have dramatically improved over the last few years. ASR systems are most often trained from ‘typical’ speech, which means that underrepresented groups don’t experience the same level of improvement. In this paper, we present and evaluate finetuning techniques to improve ASR for users with non standard speech. We focus on two types of non standard speech: speech from people with amyotrophic lateral sclerosis (ALS) and accented speech. We train personalized models that achieve 62% and 35% relative WER improvement on these two groups, bringing the absolute WER for ALS speakers, on a test set of message bank phrases, to 10% for mild dysarthria and 20% for more serious dysarthria. We show that 76% of the improvement comes from only 5 min of training data. Finetuning a particular subset of layers (with many fewer parameters) often gives better results than finetuning the entire model. This is the first step towards building state of the art ASR models for dysarthric speech Index Terms: speech recognition, personalization, accessibility View details
    Joint Cache Partition and Job Assignment on Multi-Core Processors
    WADS'13: Proceedings of the 13th international conference on Algorithms and Data Structures (2012)
    Preview abstract Multicore shared cache processors pose a challenge for designers of embedded systems who try to achieve minimal and predictable execution time of workloads consisting of several jobs. To address this challenge the cache is statically partitioned among the cores and the jobs are assigned to the cores so as to minimize the makespan. Several heuristic algorithms have been proposed that jointly decide how to partition the cache among the cores and assign the jobs. We initiate a theoretical study of this problem which we call the joint cache partition and job assignment problem. By a careful analysis of the possible cache partitions we obtain a constant approximation algorithm for this problem. For some practical special cases we obtain a 2-approximation algorithm, and show how to improve the approximation factor even further by allowing the algorithm to use additional cache. We also study possible improvements that can be obtained by allowing dynamic cache partitions and dynamic job assignments. We define a natural restriction of the well known scheduling problem on unrelated machines in which machines are ordered by “strength”. We call this restriction the ordered unrelated machines scheduling problem. We show that our joint cache partition and job assignment problem is harder than this scheduling problem. The ordered unrelated machines scheduling problem is of independent interest and we give a polynomial time algorithm for certain natural workloads. View details
    No Results Found