Jing Pan

Title: Innovations of Jing Pan in Speech Processing

Introduction

Jing Pan is an accomplished inventor based in San Jose, CA, known for his contributions to the field of speech processing. With a total of 2 patents, he has made significant strides in improving the quality of speech applications through innovative techniques.

Latest Patents

Jing Pan's latest patents include advancements in contextual feature vectors for processing speech. This invention focuses on enhancing the quality of feature vectors, which can lead to improved performance in speech applications. By modifying neural network architectures, he allocates computational resources more effectively for learning and computing feature vectors. The process involves parameterized downsampling and upsampling operations to optimize vector sequence rates.

Another notable patent is related to speech sentiment analysis using a speech sentiment classifier pretrained with pseudo sentiment labels. This system predicts sentiment labels for audio speech utterances by leveraging a classifier that has been pretrained in an unsupervised manner. By utilizing a text-trained pseudo labeler, Jing's method autogenerates pseudo sentiment labels, which are then fine-tuned using a smaller, sentiment-annotated dataset.

Career Highlights

Jing Pan is currently employed at Asapp, Inc., where he continues to develop innovative solutions in speech technology. His work has garnered attention for its potential to enhance user experiences in various applications.

Collaborations

Jing collaborates with talented individuals such as Kyu Jeong Han and Suwon Shon, contributing to a dynamic team focused on advancing speech processing technologies.

Conclusion

Jing Pan's innovative work in speech processing showcases his dedication to improving technology through his patents and collaborations. His contributions are paving the way for more effective speech applications in the future.