说谎行为识别线索的比较研究

Studies on lie-detection by western psychologists indicate that lying cues people usually hold are not in accordance with the real verbal and non-verbal behaviors that liars usually show. A cross-culture study carried out by C.F.Bond and its global research team finds that the commonest view held by people from 75 nations about lying behavior is that liars usually avert gaze, while study shows that gaze-aversion has no relation with lying. In Bond’s view, stereotype of the liar reflect more about common cross-culture values than an objective description of how liars behave. Different culture has its norms based upon which people judge whether a person is credible or not. As a nation of long Confucianism tradition, how Chinese view liars differently from people of other culture is the interest of this study. By a comparative study with that of Bond’s research, it is found that, in line with Bond’s finding, Chinese generally hold the same stereotype about liars with that of the westerners; but it seems that Chinese rely significantly less on gaze-aversion as a cue to lying, and they concern more about senders’ motivation and emotion. It is also found that confidence about their detection ability among Chinese is lower than westerners. A further study on different professions and their view about lying behaviors shows that people in law-enforcement and related professions generally hold a more accurate view toward how liars behave. Possible explanations to the above mentioned findings in view of culture differences, aspects to be improved in this study and direction of future research are discussed in the later part of the thesis.

Continuous speech recognition based on ICA and geometrical learning

We investigate the use of independent component analysis (ICA) for speech feature extraction in digits speech recognition systems. We observe that this may be true for recognition tasks based on Geometrical Learning with little training data. In contrast to image processing, phase information is not essential for digits speech recognition. We therefore propose a new scheme that shows how the phase sensitivity can be removed by using an analytical description of the ICA-adapted basis functions. Furthermore, since the basis functions are not shift invariant, we extend the method to include a frequency-based ICA stage that removes redundant time shift information. The digits speech recognition results show promising accuracy. Experiments show that the method based on ICA and Geometrical Learning outperforms HMM in a different number of training samples.

Wavelet denoising and feature extraction of seismic signal for footstep detection

Seismic sensors are widely used to detect moving target in ground sensor networks. Footstep detection is very important for security surveillance and other applications. Because of non-stationary characteristic of seismic signal and complex environment conditions, footstep detection is a very challenging problem. A novel wavelet denoising method based on singular value decomposition is used to solve these problems. The signal-to-noise ratio (SNR) of raw footstep signal is greatly improved using this strategy. The feature extraction method is also discussed after denosing procedure. Comparing, with kurtosis statistic feature, the wavelet energy feature is more promising for seismic footstep detection, especially in a long distance surveillance.

Continuous speech recognition based on ICA and geometrical learning

We investigate the use of independent component analysis (ICA) for speech feature extraction in digits speech recognition systems. We observe that this may be true for recognition tasks based on Geometrical Learning with little training data. In contrast to image processing, phase information is not essential for digits speech recognition. We therefore propose a new scheme that shows how the phase sensitivity can be removed by using an analytical description of the ICA-adapted basis functions. Furthermore, since the basis functions are not shift invariant, we extend the method to include a frequency-based ICA stage that removes redundant time shift information. The digits speech recognition results show promising accuracy. Experiments show that the method based on ICA and Geometrical Learning outperforms HMM in a different number of training samples.