Pub Date : 2014-11-20DOI: 10.1109/ICOT.2014.6954663
Zhi-Peng Fu, Yan-Ning Zhang, Hai-Yan Hou
Deep learning is the research focus in the recent years. Because of its excellent performance.it is widely used in the area of pattern recognition. Facial feature is useful for a variety of tasks, the application of deep learning in this area is also developing fast. We introduce some recent research work in this domain, and show the potential of it.
{"title":"Survey of deep learning in face recognition","authors":"Zhi-Peng Fu, Yan-Ning Zhang, Hai-Yan Hou","doi":"10.1109/ICOT.2014.6954663","DOIUrl":"https://doi.org/10.1109/ICOT.2014.6954663","url":null,"abstract":"Deep learning is the research focus in the recent years. Because of its excellent performance.it is widely used in the area of pattern recognition. Facial feature is useful for a variety of tasks, the application of deep learning in this area is also developing fast. We introduce some recent research work in this domain, and show the potential of it.","PeriodicalId":343641,"journal":{"name":"2014 International Conference on Orange Technologies","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-11-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129486427","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper, an integrated emotion regulation system (IERS) is proposed based on the regulation process model for happiness improvement. Including extracting the valuable information from user's contents on social network, the IERS analyzes users' emotion variation and semanteme reflecting to the regulation process model and aim to appropriately feedback to users. The feedback sentences are chosen from regulation corpus which is positive and motivated. The proposed IERS works at the word level and the emotional topics is classified by SVM through the corpus collected from Facebook wall, whereas feedback strategy sentences is chosen through Point-Wise Mutual Information (PMI) features. The accuracy result of seven-type emotion recognition can achieve higher than 50%. The pre- and post-experiment results are evaluated by 20 participants in one week of observation, of which the result implies the proposed system can practically improve the happiness.
{"title":"An emotional feedback system based on a regulation process model for happiness improvement","authors":"Y. Hung, Yang-Yen Ou, Ta-Wen Kuan, Chin-Hui Cheng, Jhing-Fa Wang, Jaw-Shyang Wu","doi":"10.1109/ICOT.2014.6956635","DOIUrl":"https://doi.org/10.1109/ICOT.2014.6956635","url":null,"abstract":"In this paper, an integrated emotion regulation system (IERS) is proposed based on the regulation process model for happiness improvement. Including extracting the valuable information from user's contents on social network, the IERS analyzes users' emotion variation and semanteme reflecting to the regulation process model and aim to appropriately feedback to users. The feedback sentences are chosen from regulation corpus which is positive and motivated. The proposed IERS works at the word level and the emotional topics is classified by SVM through the corpus collected from Facebook wall, whereas feedback strategy sentences is chosen through Point-Wise Mutual Information (PMI) features. The accuracy result of seven-type emotion recognition can achieve higher than 50%. The pre- and post-experiment results are evaluated by 20 participants in one week of observation, of which the result implies the proposed system can practically improve the happiness.","PeriodicalId":343641,"journal":{"name":"2014 International Conference on Orange Technologies","volume":"111 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-11-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115123877","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2014-11-20DOI: 10.1109/ICOT.2014.6956600
Shih-Pang Tseng, Bo Li, Junpeng Pan, Chia-Ju Lin
The smart house is an important issue in the orange technology. In this paper, we proposes a work which applies the Internet of things (IOT) and motion sensing on smart house, denoted by Smart House Monitor & Manager (SHMM), to improve the convenience, safety, and power-saving of the house. It is based on the Zigbee sensor connected all sensors and actuators. And we have implemented the SHMM with a smart house to demonstrate the availability.
智能住宅是橙色科技中的一个重要课题。本文提出了一项将物联网(IOT)和体感技术应用于智能家居的工作，即智能家居监控与管理(smart house Monitor & Manager, SHMM)，以提高家居的便利性、安全性和节能性。它是基于Zigbee传感器连接所有传感器和执行器。我们已经在一个智能住宅中实施了SHMM来证明它的可用性。
{"title":"An application of Internet of things with motion sensing on smart house","authors":"Shih-Pang Tseng, Bo Li, Junpeng Pan, Chia-Ju Lin","doi":"10.1109/ICOT.2014.6956600","DOIUrl":"https://doi.org/10.1109/ICOT.2014.6956600","url":null,"abstract":"The smart house is an important issue in the orange technology. In this paper, we proposes a work which applies the Internet of things (IOT) and motion sensing on smart house, denoted by Smart House Monitor & Manager (SHMM), to improve the convenience, safety, and power-saving of the house. It is based on the Zigbee sensor connected all sensors and actuators. And we have implemented the SHMM with a smart house to demonstrate the availability.","PeriodicalId":343641,"journal":{"name":"2014 International Conference on Orange Technologies","volume":"217 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-11-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121729931","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2014-11-12DOI: 10.1109/ICOT.2014.6956619
Yuh-Tyng Chen, S. Liou
Interactive online learning environment with audiovisual multimedia has improved significantly in increasing learning effectiveness. However, for hearing-impaired students who have difficulties in internet multimedia systems associated with audio and visual information require further design to overcome their limitation. To this end, the present study develops an interactive online learning environment to increase accessibility and interactivity for hearing-impaired students based on hyperlink techniques, modular concepts and online discussions. In particular, for increasing accessibility, in Course Design, to capture attention and arouse interests, 1) we provide pre-course briefing with sign language and lip language, and displaying learning outcome, for reducing cognitive loading, and 2) categorize topics and simplify course contents, for providing enough time to comprehend, moreover, we add 3)self-assessment and online forum as synchronized discussion. For increasing interactivity, in Interface Design, we provide 4) index buttons for fast understanding of course outline and 5) self selection of course, and 6) impressive visual guide to highlight important points and to help comprehension of course contents. Our empirical experiments based on 60 hearing-impaired students showed that the abovementioned five enhancement designs have significant increase their intention to use interactive online learning environment manifested in Technology Acceptance Model (TAM). The results showed that the accessibility and interactivity respectively produced significant effects on perceived ease of use and perceived usefulness, and then that the hearing-impaired students felt helpful and intended to use the interactive online learning environment as often as needed. However, it was also found that perceived ease of use did not produce direct effects on the hearing-impaired students' intention to use. The implication in design better online learning environment for particular users and contribution in developing orange technology of learning are discussed.
{"title":"Enhancing the acceptance of interactive online learning of hearing-impaired students","authors":"Yuh-Tyng Chen, S. Liou","doi":"10.1109/ICOT.2014.6956619","DOIUrl":"https://doi.org/10.1109/ICOT.2014.6956619","url":null,"abstract":"Interactive online learning environment with audiovisual multimedia has improved significantly in increasing learning effectiveness. However, for hearing-impaired students who have difficulties in internet multimedia systems associated with audio and visual information require further design to overcome their limitation. To this end, the present study develops an interactive online learning environment to increase accessibility and interactivity for hearing-impaired students based on hyperlink techniques, modular concepts and online discussions. In particular, for increasing accessibility, in Course Design, to capture attention and arouse interests, 1) we provide pre-course briefing with sign language and lip language, and displaying learning outcome, for reducing cognitive loading, and 2) categorize topics and simplify course contents, for providing enough time to comprehend, moreover, we add 3)self-assessment and online forum as synchronized discussion. For increasing interactivity, in Interface Design, we provide 4) index buttons for fast understanding of course outline and 5) self selection of course, and 6) impressive visual guide to highlight important points and to help comprehension of course contents. Our empirical experiments based on 60 hearing-impaired students showed that the abovementioned five enhancement designs have significant increase their intention to use interactive online learning environment manifested in Technology Acceptance Model (TAM). The results showed that the accessibility and interactivity respectively produced significant effects on perceived ease of use and perceived usefulness, and then that the hearing-impaired students felt helpful and intended to use the interactive online learning environment as often as needed. However, it was also found that perceived ease of use did not produce direct effects on the hearing-impaired students' intention to use. The implication in design better online learning environment for particular users and contribution in developing orange technology of learning are discussed.","PeriodicalId":343641,"journal":{"name":"2014 International Conference on Orange Technologies","volume":"104 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133211486","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2014-09-01DOI: 10.1109/ICOT.2014.6956617
Bo-Hao Su, Ping-Wen Fu, Po-Chuan Lin, Po-Yi Shih, Yuh-Chung Lin, Jhing-Fa Wang, A. Tsai
This work presents a spoken dialogue system with situation and emotion detection based on anthropomorphic learning for warming healthcare. To provide more warming feedback of the system, we combine situation and emotion detection with spoken dialogue system. Situation and emotion detection are based on lexical category using Partial-Matching Spoken Sentence Retrieval (PMSSR). Moreover, an anthropomorphic learning mechanism is proposed to improve the performance of emotion and situation detection. The mechanism based on out-of-vocabulary (OOV) detection is used to update emotion and situation database with new lexicon through interaction with user and internet. The experimental results show that the anthropomorphic learning mechanism increases the accuracy rate of situation and emotion detection by 30% and 20%, respectively.
{"title":"A spoken dialogue system with situation and emotion detection based on anthropomorphic learning for warming healthcare d","authors":"Bo-Hao Su, Ping-Wen Fu, Po-Chuan Lin, Po-Yi Shih, Yuh-Chung Lin, Jhing-Fa Wang, A. Tsai","doi":"10.1109/ICOT.2014.6956617","DOIUrl":"https://doi.org/10.1109/ICOT.2014.6956617","url":null,"abstract":"This work presents a spoken dialogue system with situation and emotion detection based on anthropomorphic learning for warming healthcare. To provide more warming feedback of the system, we combine situation and emotion detection with spoken dialogue system. Situation and emotion detection are based on lexical category using Partial-Matching Spoken Sentence Retrieval (PMSSR). Moreover, an anthropomorphic learning mechanism is proposed to improve the performance of emotion and situation detection. The mechanism based on out-of-vocabulary (OOV) detection is used to update emotion and situation database with new lexicon through interaction with user and internet. The experimental results show that the anthropomorphic learning mechanism increases the accuracy rate of situation and emotion detection by 30% and 20%, respectively.","PeriodicalId":343641,"journal":{"name":"2014 International Conference on Orange Technologies","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125514817","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this work, an efficient tumor positioning method is proposed by performing registration based segmentation from 18-FDG PET-CT scanners. At the first stage, the tumor is segmented from PET scans by region growing using the manual seeds which employs the SUV monotonous features, and then the tumor contours are transferred to corresponding CT images automatically by registration method which is based on edge preserving scale space for following radiation therapy planning. The experiments results demonstrate the efficiency of proposed method.
{"title":"Accurate tumor positioning from PET-CT by performing registration based segmentation","authors":"Dengwang Li, Xueting Liu, Jie Wang, Qinfen Wang, Jinhu Chen, Hongsheng Li, Yong Yin","doi":"10.1109/ICOT.2014.6956612","DOIUrl":"https://doi.org/10.1109/ICOT.2014.6956612","url":null,"abstract":"In this work, an efficient tumor positioning method is proposed by performing registration based segmentation from 18-FDG PET-CT scanners. At the first stage, the tumor is segmented from PET scans by region growing using the manual seeds which employs the SUV monotonous features, and then the tumor contours are transferred to corresponding CT images automatically by registration method which is based on edge preserving scale space for following radiation therapy planning. The experiments results demonstrate the efficiency of proposed method.","PeriodicalId":343641,"journal":{"name":"2014 International Conference on Orange Technologies","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131865846","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This work presents a sound recording system for elderly security with low-complexity computation. To reduce the time complexity, the proposed system utilizes the discrete wavelet transform (DWT) with only multiplication to replace the traditional modified discrete cosine transform (MDCT). Such a transform can effectively reduce the computational time of proposed system. For recording the zeroes of the wavelet coefficients, an efficient zero recording algorithm is proposed. The proposed algorithm uses different encoding mode and corresponding bit data structure to further improve compression performance. The experimental results show that the proposed system can achieve about 33-fold improvement of the computation speed compared with the baseline, with also the similar rate-distortion performance. Such results indicate that the proposed system is superior to the baseline work to further be applied in low-end sound recording device for home-care services.
{"title":"A low-complexity sound recording system for elderly security in home-care system","authors":"Chung-Hsien Chang, Po-Chuan Lin, Yu-Hao Chiu, Jhing-Fa Wang, Ta-Wen Kuan","doi":"10.1109/ICOT.2014.6956630","DOIUrl":"https://doi.org/10.1109/ICOT.2014.6956630","url":null,"abstract":"This work presents a sound recording system for elderly security with low-complexity computation. To reduce the time complexity, the proposed system utilizes the discrete wavelet transform (DWT) with only multiplication to replace the traditional modified discrete cosine transform (MDCT). Such a transform can effectively reduce the computational time of proposed system. For recording the zeroes of the wavelet coefficients, an efficient zero recording algorithm is proposed. The proposed algorithm uses different encoding mode and corresponding bit data structure to further improve compression performance. The experimental results show that the proposed system can achieve about 33-fold improvement of the computation speed compared with the baseline, with also the similar rate-distortion performance. Such results indicate that the proposed system is superior to the baseline work to further be applied in low-end sound recording device for home-care services.","PeriodicalId":343641,"journal":{"name":"2014 International Conference on Orange Technologies","volume":"77 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125490331","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper proposes a robust template based on the previously proposed ECWRT (enhanced cross word reference template) for template-based ASR, by using correlational weight adjusting method to improve robustness against elderly speech variation named CWCWRT. This work addresses two vital issues: such as outlier rejection in training set and elimination of unwanted utterances which usually happen by the elderly people. Consequently, two main steps are investigated in this paper, firstly, correlational analyzing, and secondly, weight adjusting. For experiments, the corpus is built by 30 commands in Mandarin and English collected from three elderly (age 62±3 years) and three adults (age 22±2 years) having total 30 utterances for each of them. Two types of platforms including PC and GPCE063A embedded platform are conducted, both inside test and outside test are also applied. The results show that the average recognition rate for inside testis 97% in PC simulation and 90% in the embedded platform. The outside test results are 93% and 87% in two platforms respectively. The related and previous works including cross word reference template (CWRT) and ECWRT are also compared the comparison exhibit that the proposed CWCWRT gives higher robustness and accuracy than two baselines.
本文在前人提出的增强交叉词参考模板(enhanced cross word reference template, ECWRT)的基础上，提出了一种基于模板的ASR鲁棒模板，采用相关权值调整方法提高了对老年人语音变异的鲁棒性，称为CWCWRT。这项工作解决了两个至关重要的问题:例如训练集的异常值拒绝和消除通常发生在老年人身上的不想要的话语。因此，本文主要研究了两个步骤，首先是相关性分析，其次是权重调整。在实验中，语料库由3名老年人(62±3岁)和3名成年人(22±2岁)的30个普通话和英语命令组成，每个命令各30个话语。在PC和GPCE063A嵌入式平台两种平台上进行了内部测试和外部测试。结果表明，PC仿真对内胆的平均识别率为97%，嵌入式平台的平均识别率为90%。在两个平台上的外部测试结果分别为93%和87%。本文还比较了交叉词参考模板(cross word reference template, CWRT)和ECWRT的相关研究成果，结果表明本文提出的CWCWRT具有更高的鲁棒性和准确性。
{"title":"A high-accuracy ASR technique based on correlational weight analysis for elderly users","authors":"Chih-Hung Chou, Ta-Wen Kuan, Po-Chuan Lin, Jhing-Fa Wang, Yi-Jhong Wu","doi":"10.1109/ICOT.2014.6956631","DOIUrl":"https://doi.org/10.1109/ICOT.2014.6956631","url":null,"abstract":"This paper proposes a robust template based on the previously proposed ECWRT (enhanced cross word reference template) for template-based ASR, by using correlational weight adjusting method to improve robustness against elderly speech variation named CWCWRT. This work addresses two vital issues: such as outlier rejection in training set and elimination of unwanted utterances which usually happen by the elderly people. Consequently, two main steps are investigated in this paper, firstly, correlational analyzing, and secondly, weight adjusting. For experiments, the corpus is built by 30 commands in Mandarin and English collected from three elderly (age 62±3 years) and three adults (age 22±2 years) having total 30 utterances for each of them. Two types of platforms including PC and GPCE063A embedded platform are conducted, both inside test and outside test are also applied. The results show that the average recognition rate for inside testis 97% in PC simulation and 90% in the embedded platform. The outside test results are 93% and 87% in two platforms respectively. The related and previous works including cross word reference template (CWRT) and ECWRT are also compared the comparison exhibit that the proposed CWCWRT gives higher robustness and accuracy than two baselines.","PeriodicalId":343641,"journal":{"name":"2014 International Conference on Orange Technologies","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122874798","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2014-09-01DOI: 10.1109/ICOT.2014.6956623
Min Shih, Yan-Yu Lin, Yu-Shan Lin, Chang-Hong Lin, Ming-Yen Chen, Jyun-Hong Li, Chih-Wei Su, Jia-Ching Wang
We combine sound activity detection, sound enhancement, and direction of arrival (DOA) estimation system to an integrated system. We use sound activity detection for finding frames which are sound-dominated, then pass the signal to the subspace based sound enhancement for denoising. Finally, the denoised signal will be input to DOA detection system for sound tracking. The generalized cross-correlation and phase transformation (GCC-PHAT) based time difference of arrival (TDOA) estimation can detect the TDOA of the sound signal and then calculate the respondent DOA.
{"title":"System implementation of robust time difference of arrival estimation","authors":"Min Shih, Yan-Yu Lin, Yu-Shan Lin, Chang-Hong Lin, Ming-Yen Chen, Jyun-Hong Li, Chih-Wei Su, Jia-Ching Wang","doi":"10.1109/ICOT.2014.6956623","DOIUrl":"https://doi.org/10.1109/ICOT.2014.6956623","url":null,"abstract":"We combine sound activity detection, sound enhancement, and direction of arrival (DOA) estimation system to an integrated system. We use sound activity detection for finding frames which are sound-dominated, then pass the signal to the subspace based sound enhancement for denoising. Finally, the denoised signal will be input to DOA detection system for sound tracking. The generalized cross-correlation and phase transformation (GCC-PHAT) based time difference of arrival (TDOA) estimation can detect the TDOA of the sound signal and then calculate the respondent DOA.","PeriodicalId":343641,"journal":{"name":"2014 International Conference on Orange Technologies","volume":"2000 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123550327","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：