Speech Synthesis Engineer
AI Rudder is a business solution and service provider for small and medium businesses as well as enterprises.
Leveraging AI technologies of natural languages, AI Rudder brings value to business procedures by lowering the cost, mitigating the risk, and increasing the profit. As a fast-growing startup, AI Rudder will make sure everyone will practice, learn and grow together. At this stage, AI Rudder focuses on AI call center solution for financial industry in Southeast Asia and India subcontinent, where great progress has been made. That is why AI Rudder needs more talents such as you to join and create a new horizon that utilizes technology to change the world.
Responsibilities
- Front-end text analysis and normalization processing;
- Analysis techniques such as rhythm and duration;
- The landing of hybrid splicing method based on neural network;
- End-to-end generative speech synthesis algorithm and engineering optimization;
Minimum Qualifications
- Strong anti-stress ability, adapt to the rhythm of startup companies, hoping to win the future;
- Have knowledge and professional experience in speech synthesis R&D;
- Master degree or above, major in signal processing, computer, electronic information, automation, pattern recognition, etc.;
- Knowledge background related to speech recognition, speech signal processing, pattern recognition, deep learning;
- Development and debugging experience in Linux environment, proficient programming skills (C/C++), familiar with a scripting language (Perl/python), will use Shell programming;
- Proficient in one or more of community open source tools such as HTS, Merlin, TensorFlow, etc.;
- Innovative and critical thinking;
- Good communication skills;
- Responsible, passionate, serious and result-oriented;
- A degree in engineering or technology.
Preferred Qualifications
- Those who published papers in relevant international conferences or mainstream journals are preferred (ICASSP, Interspeech), and those with experience in participating in the Blizzard competition are preferred;
- Familiar with the latest end-to-end (such as Tacotron*, deepvoice*, waveglow, etc.) synthesis technology is preferred.