Current Activity
I'm a Mater student at the Digital Audio and Image Signal Processing Group (DAISPG) of the UESTC, Chengdu, Sichuan province,China, under the supervision of Gan Tao Associate Professor.
My main field of interest is Audio/Speech Information Processing, with a focus on content-based MIR and audio and music processing. My research interests include Speech Recognition, Music Vocal/Nonvocal Segment, F0 estimation, Music Information Retrieval(MIR),Speaker Diarization, Data Mining.
Project
I have finish most of it and now mainly study in accompaniment reduction which is crucial to this project.
Platform:Windows;Language:Matlab.
an unknown number of speakers.This project is mainly used in broadcast news.
Platform:Linux;Language:C.
I have finished this project.The approach which I have used is two-step retrieval mainly based on distance.The first step is based on cosine distance to roughly select clips which are similar to the short audio segment (the segment we'll search);The second step mainly uses eulicdean distance to find the correct segments.Finally,we should delete the segments adjacently.OK,we got this problem solved.
Platform:Windows;Language:C++.
Teaching
I'm recently instructing three undergraduate students with their bachelor thesis.The titles are:
Software
2. Platform
Android 1.6 or above
3. Download
- Vocal Detection of Popular Music
I have finish most of it and now mainly study in accompaniment reduction which is crucial to this project.
Platform:Windows;Language:Matlab.
- Speaker Diarization
an unknown number of speakers.This project is mainly used in broadcast news.
Platform:Linux;Language:C.
- Audio Clips Retrieval
I have finished this project.The approach which I have used is two-step retrieval mainly based on distance.The first step is based on cosine distance to roughly select clips which are similar to the short audio segment (the segment we'll search);The second step mainly uses eulicdean distance to find the correct segments.Finally,we should delete the segments adjacently.OK,we got this problem solved.
Platform:Windows;Language:C++.
- Anchor speakers tracking in broadcast news
- Web-based mandarin broadcast news retrieval
Teaching
I'm recently instructing three undergraduate students with their bachelor thesis.The titles are:
- Music vocal/nonvocal segment
- Study on speaker recognition
- Speech emotion recognition
Software
- DialectSpeaker V1.0
- Introduction
2. Platform
Android 1.6 or above
3. Download
dialectspeaker_v1.0.apk | |
File Size: | 644 kb |
File Type: | apk |