Shih-Hung Yang

Prof. Shih-Hung Yang

National Cheng Kung University
orcid

Research Focus & Metrics

Academic Career Length

18Years (2009-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 June 2012 in Neurocomputing
Verified
Journal Article
Closed access
SYShih-Hung Yang
Shih-Hung Yang
YCYon-Ping Chen
Yon-Ping Chen

7 December 2017 in Symmetry
Verified
Journal Article
Open access
PDF file
SWShun-Yi Wang
Shun-Yi Wang
SYShih-Hung Yang
Shih-Hung Yang
YCYon-Ping Chen
Yon-Ping Chen
JHJyun-We Huang
Jyun-We Huang
Face recognition systems have been widely adopted for user authentication in security systems due to their simplicity and effectiveness. However, spoofing attacks, including printed photos, displayed photos, and replayed video attacks, are critical challenges to authentication, and these spoofing attacks allow malicious invaders to gain access to the system. This paper proposes two novel features for face liveness detection systems to protect against printed photo attacks and replayed attacks for biometric authentication systems. The first feature obtains the texture difference between red and green channels of face images inspired by the observation that skin blood flow in the face has properties that enable distinction between live and spoofing face images. The second feature estimates the color distribution in the local regions of face images, instead of whole images, because image quality might be more discriminative in small areas of face images. These two features are concatenated together, along with a multi-scale local binary pattern feature, and a support vector machine classifier is trained to discriminate between live and spoofing face images. The experimental results show that the performance of the proposed method for face spoof detection is promising when compared with that of previously published methods. Furthermore, the proposed system can be implemented in real time, which is valuable for mobile applications.

3 December 2021 in Biosensors
Verified
Journal Article
Open access
PDF file
SYShih-Hung Yang
Shih-Hung Yang
CKChia-Lin Koh
Chia-Lin Koh
CHChun-Hang Hsu
Chun-Hang Hsu
PCPo-Chuan Chen
Po-Chuan Chen
JCJia-Wei Chen
Jia-Wei Chen
YCYou-Yin Chen
You-Yin Chen
Effective bilateral hand training is desired in rehabilitation programs to restore hand function for people with unilateral hemiplegia, so that they can perform daily activities independently. However, owing to limited human resources, the hand function training available in current clinical settings is significantly less than the adequate amount needed to drive optimal neural reorganization. In this study, we designed a lightweight and portable hand exoskeleton with a hand-sensing glove for bilateral hand training and home-based rehabilitation. The hand-sensing glove measures the hand movement of the less-affected hand using a flex sensor. Thereafter, the affected hand is driven by the hand exoskeleton using the measured hand movements. Compared with the existing hand exoskeletons, our hand exoskeleton improves the flexible mechanism for the back of the hand for better wearing experience and the thumb mechanism to make the pinch gesture possible. We designed a virtual reality game to increase the willingness of repeated movement practice for rehabilitation. Our system not only facilitates bilateral hand training but also assists in activities of daily living. This system could be beneficial for patients with hemiplegia for starting correct and sufficient hand function training in the early stages to optimize their recovery.

12 December 2017 in Frontiers in Neuroscience
Verified
Journal Article
Open access
SLSheng-Huang Lin
Sheng-Huang Lin
HLHsin-Yi Lai
Hsin-Yi Lai
YLYu-Chun Lo
Yu-Chun Lo
CCChin Chou
Chin Chou
YCYi-Ting Chou
Yi-Ting Chou
YCYou-Yin Chen
You-Yin Chen
Deep brain stimulation (DBS) surgery of the subthalamic nucleus (STN) under general anesthesia (GA) had been used in Parkinson's disease (PD) patients who are unable tolerate awake surgery. The effect of anesthetics on intraoperative microelectrode recording (MER) remains unclear. Understanding the effect of anesthetics on MER is important in performing STN DBS surgery with general anesthesia. In this study, we retrospectively performed qualitive and quantitative analysis of STN MER in PD patients received STN DBS with controlled desflurane anesthesia or LA and compared their clinical outcome. From January 2005 to March 2006, 19 consecutive PD patients received bilateral STN DBS surgery in Hualien Tzu-Chi hospital under either desflurane GA (n = 10) or LA (n = 9). We used spike analysis (frequency and modified burst index [MBI]) and the Hilbert transform to obtain signal power measurements for background and spikes, and compared the characterizations of intraoperative microelectrode signals between the two groups. Additionally, STN firing pattern characteristics were determined using a combined approach based on the autocorrelogram and power spectral analysis, which was employed to investigate differences in the oscillatory activities between the groups. Clinical outcomes were assessed using the Unified Parkinson's Disease Rating Scale (UPDRS) before and after surgery. The results revealed burst firing was observed in both groups. The firing frequencies were greater in the LA group and MBI was comparable in both groups. Both the background and spikes were of significantly greater power in the LA group. The power spectra of the autocorrelograms were significantly higher in the GA group between 4 and 8 Hz. Clinical outcomes based on the UPDRS were comparable in both groups before and after DBS surgery. Under controlled light desflurane GA, burst features of the neuronal firing patterns are preserved in the STN, but power is reduced. Enhanced low-frequency (4–8 Hz) oscillations in the MERs for the GA group could be a characteristic signature of desflurane's effect on neurons in the STN.

1 May 2017 in Brain Stimulation
Verified
Journal Article
Closed access
CWChing-Fu Wang
Ching-Fu Wang
SYShih-Hung Yang
Shih-Hung Yang
SLSheng-Huang Lin
Sheng-Huang Lin
PCPo-Chuan Chen
Po-Chuan Chen
YLYu-Chun Lo
Yu-Chun Lo
YCYou-Yin Chen
You-Yin Chen

24 September 2021 in Sensors
Verified
Journal Article
Open access
PDF file
SYShih-Hung Yang
Shih-Hung Yang
JHJyun-We Huang
Jyun-We Huang
CHChun-Jui Huang
Chun-Jui Huang
PCPo-Hsiung Chiu
Po-Hsiung Chiu
HLHSIN-YI LAI
HSIN-YI LAI
YCYou-Yin Chen
You-Yin Chen
Intracortical brain–computer interfaces (iBCIs) translate neural activity into control commands, thereby allowing paralyzed persons to control devices via their brain signals. Recurrent neural networks (RNNs) are widely used as neural decoders because they can learn neural response dynamics from continuous neural activity. Nevertheless, excessively long or short input neural activity for an RNN may decrease its decoding performance. Based on the temporal attention module exploiting relations in features over time, we propose a temporal attention-aware timestep selection (TTS) method that improves the interpretability of the salience of each timestep in an input neural activity. Furthermore, TTS determines the appropriate input neural activity length for accurate neural decoding. Experimental results show that the proposed TTS efficiently selects 28 essential timesteps for RNN-based neural decoders, outperforming state-of-the-art neural decoders on two nonhuman primate datasets (R2=0.76±0.05 for monkey Indy and CC=0.91±0.01 for monkey N). In addition, it reduces the computation time for offline training (reducing 5–12%) and online prediction (reducing 16–18%). When visualizing the attention mechanism in TTS, the preparatory neural activity is consecutively highlighted during arm movement, and the most recent neural activity is highlighted during the resting state in nonhuman primates. Selecting only a few essential timesteps for an RNN-based neural decoder provides sufficient decoding performance and requires only a short computation time.