Detecting sports fatigue from speech by support vector machine

Fatigue is a complex physiological phenomena which is a kind of human body's natural response and self-regulation for protection. Detection of fatigue is becoming indispensable for its positive significance in scientific physical training. Recently, many researchers from both speech signal area and machine learning area have already shown that automatically fatigue detection from speech can carry out, but there is still plenty of room for the improvement of the recognition accuracy. The key to raise the accuracy in voice-based fatigue detection is precise phonetic identification and alignment. Therefore, this paper proposes a method for detecting sports fatigue which is based on feature extraction and machine learning system - support vector machine (SVM). In order to establish a comprehensive identification system, speech samples are trained as speech sources at different times. Experimental results state the feasibility and effectiveness of this method we put forward. What's more, Receiver Operating Characteristic Curves (ROC Curves) are used to double check the results, so that the application of sports fatigue detection is ensured.