Scholar
Viet Anh Trinh
Google Scholar ID: Nss6O2oAAAAJ
Nvidia
Citations & Impact
Citations
159
H-index
7
i10-index
6
Publications
20
Co-authors
0
Contact
Publications
Resume (English only)
Academic Achievements
- Published 3 conference papers (Interspeech, ICASSP) and 1 journal article (TASLP). Contributions include proposing an adversarial-based approach for predicting 2-D importance maps, developing a novel evaluation metric for assessing importance maps in structured prediction tasks, and creating a data augmentation technique based on importance maps.
Research Experience
- Conducts research on developing new methods to predict and evaluate importance maps in Automatic Speech Recognition (ASR) systems, and using these maps to improve the model's performance under noisy conditions.
Education
- PhD - Department of Computer Science, CUNY Graduate Center; Advisor: Professor Michael Mandel
Background
- Currently a final-year PhD student in the Department of Computer Science at CUNY Graduate Center, focusing on Machine Learning and Speech Processing. Specifically, his research is about finding time-frequency regions in spectrograms that ASR pays attention to, known as audible importance/attention maps, which can enhance model's interpretability and performance.
Miscellany
- Contact: Email - vtrinh@gradcenter.cuny.edu and anhtv1@gmail.com; Skype: tvanh512
Co-authors