职位描述
1、负责开发新一代Kaldi(https://github.com/k2-fsa)。该项目为PyTorch添加了对加权有限状态机的支持,旨在用于语音识别和相关应用中;
2、基于新一代Kaldi,参与新型语音识别模型的研究,与Daniel Povey合作发表论文;
Job responsibilities:
1.This position involves developing the open source project "Next-Gen Kaldi" (https://github.com/k2-fsa). This project adds efficient support for Finite State Automata (FSAs) to PyTorch, and is intended for use in speech recognition and related applications.
2. Based on the next generation Kaldi, participate in the research of new speech recognition models, and publish papers in collaboration with Daniel Povey.
3. Cooperate with Xiaomi's speech team to create voice recognition algorithms for Xiaomi's devices (including mobile phones, smart speakers, smart wearable and other AIoT devices).
职位要求
任职要求:
3、熟悉TensorFlow或PyTorch等深度学习框架; 加分项:
(3)熟悉Kaldi,了解ESPnet等语音识别工具包,了解wav2letter、RNN-Transducer等语音识别模型;
4、拥有 PyTorch/ TensorFlow核心代码开发经验;
5、具备开源项目开发经验;
6、优秀的数理基础,能够快速理解掌握语音识别领域的数值计算法。
Job requirements:
1. Master's degree or above, CS/EE/mathematics related majors;
2. Excellent software engineering literacy, proficient in Python, C/C++ programming;
3. Familiar with deep learning frameworks such as TensorFlow or PyTorch; Also(these are bonus items):
(1)Good English communication skills;
(2) Proficient in computer architecture, with rich experience in parallel programming; familiar with CUDA programming;
(3)Familiar with Kaldi, understand ESPnet and other speech recognition toolkits, understand wav2letter, RNN-Transducer and other speech recognition models;
4. Experience with PyTorch/TensorFlow core code development;
5. Experience with open source project developmen;
6. Excellent mathematical foundation, familiar with machine learning.
还没有评论呢,快来抢沙发~