西瓜树-就业加油站-职业教育行业垂直平台-西瓜树

当前位置：西瓜树首页 > 找工作 > 职位列表 > 职位详情

语音合成工程师 50-80K·13薪

收藏在线沟通 投递简历

完善在线简历 上传附件简历

职位描述公司简介公司地址

职位描述

Do you want to change the way the world interacts with computers? Do you want to be part of a team that pushes the Natural User Experience to the next level? Do you dream that one day

our world will be populated by robots that will help to do our jobs? Do you want to challenge yourself by innovating in an area that s new to Microsoft yet s an important strategic bet? Do you want to make Microsoft products not only accessible

but highly functional to all the users on the world? As both computational horsepower and storage capacity reach unprecedented levels

humans are getting closer and closer to that dream of the natural user interface. Each day we are stepping closer toward being able to interact with computers the same way we interact with another human being.

The Azure text to speech’s mission s to empower every person and every organization on the planet to have human like

diverse and delightful AI voices! The TTS platform (runtime

model and services) we built has been widely used by many Microsoft products and Azure customers in voice assistant

read aloud and accessibility scenarios. We are looking for a motivated

self-driven software development engineer / applied scientist to drive the development of neural text to speech language development for key speech customers
Responsibilities

Advance the state of the art of speech technology through end-to-end modelling.
Improve the speech synthesis quality and performance in terms of naturalness

expressiveness

accuracy for production.
Debug voice quality issues in new languages
Build high quality Neural TTS using low resource data and joint learning with ASR.
Collaborate with remote teams to deliver high quality products.

Qualifications
3+ years of experience in speech synthesis or speech recognition. (required)
PhD/MS Degree in speech synthesis

or equivalent experience (list what s equivalent). (required)
Experience in end-to-end speech modelling (transformer speech etc) (required)
Ability to write deep learning code and implement state of art paper ideas. (preferred)
Understanding of neural acoustic modelling or vocoder. (preferred)
Experience in speech recognition. (preferred)

公司简介

微软，是一家美国跨国科技公司，也是世界PC（Personal Computer，个人计算机）软件开发的先导，由比尔·盖茨与保罗·艾伦创办于1975年，公司总部设立在华盛顿州的雷德蒙德（Redmond，邻近西雅图）。以研发、制造、授权和提供广泛的电脑软件服务业务为主。最为著名和畅销的产品为Microsoft Windows操作系统和Microsoft Office系列软件，目前是全球最大的电脑软件提供商。