Hi, my name is Reuben Morais
I'm a software engineer with 12 years of experience. Currently I'm working at voize on on-device ASR + NLP for professional documentation. My past focuses include, from most recent to oldest:
- Leading implementation and operation of Coqui Studio, a Web-based text-to-speech synthesis studio using Coqui's in-house TTS models.
- Building and maintaining Mozilla DeepSpeech and Coqui STT: deep learning based speech-to-text model training, optimization and on-device deployment.
- Operating system and browser development at Mozilla on the Firefox OS project.
I was a co-founder and CTO at Coqui. You can find some of my open source work on GitHub. My email address is me @ this domain.
-- reuben
Articles:
- DeepSpeech 0.9 - mature, fast, polished, ubiquitous
- Common Voice: A Massively-Multilingual Speech Corpus
- DeepSpeech 0.6: Mozilla's Speech-to-Text Engine Gets Fast, Lean, and Ubiquitous
- Streaming RNNs in TensorFlow
- A Journey to <10% Word Error Rate
Citations:
- Toward Automated Classroom Observation: Multimodal Machine Learning to Estimate CLASS Positive Climate and Negative Climate
- DashCam Pay: A System for In-vehicle Payments Using Face and Voice
- Artie Bias Corpus: An Open Dataset for Detecting Demographic Bias in Speech Applications
- Training Speech Recognition Models on HPC Infrastructure
- Learning to detect named entities in bilingual code-mixed open speech corpora
- About my Qt times, and a Qt for Python voice assistant
- How to build a voice assistant with open source Rasa and Mozilla tools