Feiping Li

Title: Feiping Li: Innovator in Voice Activity Detection Technology

Introduction

Feiping Li is a prominent inventor based in Santa Clara, CA, known for his contributions to voice activity detection technology. With a focus on hybrid machine learning techniques, he has developed innovative solutions that enhance the accuracy of speech detection in various applications.

Latest Patents

Feiping Li holds a patent titled "Hybrid learning-based and statistical processing techniques for voice activity detection." This patent describes a hybrid machine learning-based and digital signal processing (DSP) statistical post-processing technique for detecting voice activity. The technique utilizes a deep neural network (DNN) model with a small context window to estimate the probability of speech by frames. The DSP statistical post-processing stage operates on the frame-based speech probabilities from the DNN model to smooth the probabilities and reduce transitions between speech and non-speech states. This hybrid technique estimates soft decisions on detected speech in each frame based on smoothed probabilities, generates hard decisions using a threshold, detects complete utterances that may include brief pauses, and estimates the endpoint of the utterance. Additionally, the technique incorporates a target directional probability estimator to assess the direction of the speech source, enhancing the accuracy of voice activity detection.

Career Highlights

Feiping Li is currently employed at Apple Inc., where he continues to innovate in the field of voice technology. His work has significantly contributed to advancements in how devices recognize and process human speech, making interactions more seamless and efficient.

Collaborations

Feiping has collaborated with notable colleagues, including Ramin Pishehvar and Ante Jukic, who share his passion for advancing technology in voice recognition and processing.

Conclusion

Feiping Li's innovative work in voice activity detection exemplifies the impact of hybrid machine learning techniques in modern technology. His contributions continue to shape the future of voice recognition systems, enhancing user experiences across various platforms.