Liang Gu

Title: The Innovations of Liang Gu

Introduction

Liang Gu is an accomplished inventor based in Elmsford, NY (US). He has made significant contributions to the field of speech recognition technology. With a total of 2 patents, his work has advanced the understanding and implementation of perceptual harmonic cepstral coefficients.

Latest Patents

Gu's latest patents focus on the use of perceptual harmonic cepstral coefficients as a front-end for speech recognition. His innovative approach involves pitch estimation and classification into voiced, unvoiced, and transitional speech using a spectro-temporal auto-correlation technique. A peak picking formula is employed to enhance accuracy. Additionally, a weighting function is applied to the power spectrum, followed by mel-scaled band-pass filtering of the harmonics weighted power spectrum. The log-energy of the filter's output is then transformed using discrete cosine transformation to produce cepstral coefficients. To further refine the process, a within-filter cubic-root amplitude compression is applied, which reduces amplitude variation while maintaining gain invariance properties.

Career Highlights

Liang Gu is affiliated with the University of California, where he continues to push the boundaries of research in speech recognition. His work has garnered attention for its innovative methodologies and practical applications in technology.

Collaborations

Gu collaborates with Kenneth Rose, a fellow researcher, to explore advancements in their field. Their partnership has led to significant developments in speech recognition technologies.

Conclusion

Liang Gu's contributions to speech recognition through his innovative patents and collaborations highlight his role as a leading inventor in this field. His work continues to influence the future of technology in meaningful ways.