Rewon Child

Title: Rewon Child: Innovator in Speech Recognition Technologies

Introduction

Rewon Child is a prominent inventor based in San Francisco, CA, known for his contributions to the field of speech recognition technologies. With a focus on reducing bias in speech models and enhancing keyword spotting systems, he has made significant strides in the industry. His innovative work has led to the development of two notable patents.

Latest Patents

Rewon Child's latest patents include "Systems and methods for principled bias reduction in production speech models." This patent describes systems and methods to identify and address sources of bias in an end-to-end speech model. The model may utilize a recurrent neural network with two 2D-convolutional input layers, followed by multiple bidirectional recurrent layers and one fully connected layer before a softmax layer. The network is trained end-to-end using the CTC loss function to directly predict sequences of characters from log spectrograms of audio. This approach aims to remove unwanted bias induced by using purely forward-only recurrences in a deployed model.

Another significant patent is "Convolutional recurrent neural networks for small-footprint keyword spotting." This patent outlines systems and methods for creating and using Convolutional Recurrent Neural Networks (CRNNs) for small-footprint keyword spotting systems. The design combines the strengths of convolutional layers to utilize data structure in time and frequency domains with recurrent layers to leverage context for the entire processed frame. The architecture parameters were examined to determine preferred model embodiments, balancing performance and model size. The CRNN model demonstrated high accuracy and robust performance while using only approximately 230 k parameters, yielding acceptably low latency.

Career Highlights

Rewon Child is currently employed at Baidu USA LLC, where he continues to innovate in the field of speech recognition. His work has garnered attention for its practical applications and contributions to advancing technology in this area.

Collaborations

Rewon has collaborated with notable colleagues, including Adam Coates and Christopher Fougner, who share his passion for advancing speech recognition technologies.

Conclusion

Rewon Child's innovative patents and contributions to speech recognition technologies highlight his role as a leading inventor in the field. His work continues to influence the development of more efficient and unbiased speech models.