Location History:
- San Francisco, CA (US) (2020)
- Mountain View, CA (US) (2020)
Company Filing History:
Years Active: 2020
Title: Innovations of Ajay Kannan in Neural Text-to-Speech Technology
Introduction
Ajay Kannan is a prominent inventor based in San Francisco, CA. He has made significant contributions to the field of neural text-to-speech (TTS) technology. With a total of 2 patents, his work focuses on enhancing the naturalness and efficiency of speech synthesis systems.
Latest Patents
One of Ajay Kannan's latest patents is titled "Systems and methods for neural text-to-speech using convolutional sequence learning." This patent describes a fully-convolutional attention-based neural TTS system known as Deep Voice 3. The system matches state-of-the-art neural speech synthesis systems in naturalness while training ten times faster. Deep Voice 3 has been scaled to unprecedented data set sizes for TTS, utilizing over eight hundred hours of audio from more than two thousand speakers. The patent also addresses common error modes in attention-based speech synthesis networks and presents methods to scale inference to ten million queries per day on a single-GPU server.
Another significant patent is "Systems and methods for principled bias reduction in production speech models." This invention identifies and addresses sources of bias in end-to-end speech models. The model may include a recurrent neural network with two 2D-convolutional input layers, followed by multiple bidirectional recurrent layers. The network is trained end-to-end using the CTC loss function to predict sequences of characters from log spectrograms of audio. This approach helps to mitigate unwanted bias in deployed models.
Career Highlights
Ajay Kannan currently works at Baidu USA LLC, where he continues to innovate in the field of speech technology. His work has been instrumental in advancing the capabilities of neural TTS systems.
Collaborations
Ajay collaborates with talented individuals such as Sercan Omer Arik and Wei Ping, contributing to the development of cutting-edge technologies in speech synthesis.
Conclusion
Ajay Kannan's contributions to neural text-to-speech technology demonstrate his commitment to innovation and excellence in the field. His patents reflect a deep understanding of the challenges in speech synthesis and offer solutions that enhance both performance and fairness in speech models.