A5068799580- Speech and Audio Processing
- Speech Recognition and Synthesis
- Voice and Speech Disorders
- Music and Audio Processing
- Animal Vocal Communication and Behavior
- Sports injuries and prevention
- Embedded Systems Design Techniques
- Robotic Path Planning Algorithms
- Industrial Vision Systems and Defect Detection
- Advanced Neural Network Applications
- Medical Research and Treatments
- Advanced Machining and Optimization Techniques
- VLSI and FPGA Design Techniques
- Metaheuristic Optimization Algorithms Research
- Respiratory and Cough-Related Research
- Vehicle License Plate Recognition
- IoT and GPS-based Vehicle Safety Systems
- Real-Time Systems Scheduling
- Phonetics and Phonology Research
- Educational Technology and Pedagogy
- Smart Parking Systems Research
- Oral and Craniofacial Lesions
- Music Technology and Sound Studies
- Dysphagia Assessment and Management
- Sports and Physical Education Research
Chinese University of Hong Kong, Shenzhen
2024
Tianjin University
2018-2023
RELX Group (Netherlands)
2023
Gannan Medical University
2014
Individuals, such as voice-related professionals, elderly people and smokers, are increasingly suffering from voice disorder, which implies the importance of pathological repair. Previous work on repair only concerned about sustained vowel /a/, but multiple vowels is still challenging due to unstable extraction pitch unsatisfactory reconstruction formant. In this paper, a based Line Spectrum Pair feature for disorder proposed, broadened research subjects single /a/ /i/ /u/ achieved these...
At present, pathological voice recognition is mainly based on the classification of voice. However, almost all researches are single vowel \a\ samples, but few multi-vowels. In addition, current multi-vowels for normal voices, which unsuitable speech and simultaneously. This paper concentrates developing an accurate robust feature called enhanced-bark line spectrum pair (E-BLSP) to detect classify We explore impact E-BLSP performance propose effective method combination three features...
Hardware/software partitioning plays an important role in the co-design system of software and hardware. It can improve performance embedded to a great degree. Multi-objective hardware/software aims optimize from multi-aspects simultaneously. In recent years, more heuristic algorithms are utilized solve multi-objective problems. this paper, we apply firework algorithm (FWA) problem partitioning. The sorting method for solutions is described detail. calculation explosion amplitude modified...
This paper introduces DysArinVox, a new pathological speech corpus in Chinese. It included 173 participants from 27 healthy individuals and 146 voice disorders, whose various types severities of vocal impairments as diagnosed by pathology experts via auditory perceptual evaluations laryngoscopic imagery. DysArinVox is designed to provide high-quality Chinese resource for AI-driven diagnostics prognostics. To ensure the efficiency collection, we meticulously crafted recording scripts...
AI-driven severity assessment techniques for dysarthric disorders show promise in aiding speech-language pathologists with diagnostics and therapeutic follow-ups patients. Existing solutions generally focus on the average intelligibility hoarseness of individual speaker's speech (i.e., speaker-level classification). This potentially ignores slight variations pronunciation attributed to disorders, e.g., /t/ /d/. To address this issue, we rethink inherent differences dysarthria speech, propose...
For a long time, the school's football teaching has had it with basic knowledge, technology, skills and monotonous technical method, so, is easy to make students feel bored lose interest confidence in learning.In teaching, taking game as an assisted instruction of not only conforms modern thought--teaching through lively activities, but also helps stimulate students' learning motivation, which can improve level consciousness more comprehensively train spirit collectivism work solidly...
Football constantly has a great popularity among students, so, the development of campus football game grows vigorously.But dangerousness still cannot be avoided in exercise, so injury accident often occurs class, therefore, it necessary to carry out safety education school teaching.This paper analyzes causes accidents teaching, and puts forward corresponding countermeasures, as ensure students' provide beneficial reference improve quality teaching.
Acoustic Scene Classification (ASC) aims to obtain the sound environment by analyzing audio signals. Due low complexity and acquisition cost of signals, ASC has enormous potential in various applications, such as audio-based surveillance, smart cities/homes, robotics. Recently, methods have been proposed for achieved good performance. However, when they are used address complex problems, most them suffer from low-performance problem. In this paper, we propose use hierarchical classification...
Vocoder-based speech synthesis has become a promising technique to accommodate the demands of high-quality analysis, manipulation, and synthesis. However, most existing works focus on how synthesize normal human voice with high signal-to-noise ratio, neglecting individuals' pathological disorder in interaction. In this work, we propose non-linear repair vocoder for vowels sentences, which takes as input generates repaired speech. Our approach is specifically designed enhance quality...