Xianglilan Zhang
Papers
1
Total Citations
16
H-Index
1
About
Xianglilan Zhang is a researcher whose work sits at the intersection of speech recognition, privacy-preserving technology, and language processing. Her most cited paper, "One-against-All Weighted Dynamic Time Warping for Language-Independent and Speaker-Dependent Speech Recognition in Adverse Conditions" (2014, 16 citations), addresses a critical challenge: enabling accurate speech recognition for rare words and non-English names while protecting personal privacy. This work proposes a novel approach that combines language-independent processing with lightweight speaker-dependent adaptation, using weighted dynamic time warping to improve robustness in noisy environments. The method is particularly valuable for applications where training data is scarce and privacy concerns limit data collection. Zhang's contribution lies in demonstrating that effective ASR can be achieved without requiring extensive, privacy-invasive training datasets, making speech technology more accessible and ethically sound. Her research has implications for voice-activated systems, assistive technologies, and multilingual communication tools. By focusing on the intersection of privacy and performance, Zhang has carved out a niche that addresses both technical and societal challenges in speech processing.
Research Focus
Key Achievements
Top Papers
- 1