Pub Date : 2016-10-01DOI: 10.1109/GCCE.2016.7800355
Hailong Liu, T. Taniguchi, Kazuhito Takenaka, Yusuke Tanaka, T. Bando
Analyzing driving behavior data is essential for developing driver assistance systems. Statistical segmentation is one of the important methods to realize the analysis. Driving behavior data actually include undesirable defects caused by external environment and sensor failures. Defects in the data cause a huge negative effect on the segmentation. In this paper, we showed that a feature extraction method based on a deep sparse autoencoder with fixed point (DSAE-FP) could reduce the negative effect of defective data in a driving behavior segmentation task. In the experiments, we used sticky hierarchical Dirichlet process hidden Markov model to segment the driving behavior. We compared the segmentation results using hidden features extracted by DSAE-FP and other comparative methods. Experimental results showed that segmentation results of non-defective dataset and defective dataset turned out most similar when DSAE-FP was used.
{"title":"Reducing the negative effect of defective data on driving behavior segmentation via a deep sparse autoencoder","authors":"Hailong Liu, T. Taniguchi, Kazuhito Takenaka, Yusuke Tanaka, T. Bando","doi":"10.1109/GCCE.2016.7800355","DOIUrl":"https://doi.org/10.1109/GCCE.2016.7800355","url":null,"abstract":"Analyzing driving behavior data is essential for developing driver assistance systems. Statistical segmentation is one of the important methods to realize the analysis. Driving behavior data actually include undesirable defects caused by external environment and sensor failures. Defects in the data cause a huge negative effect on the segmentation. In this paper, we showed that a feature extraction method based on a deep sparse autoencoder with fixed point (DSAE-FP) could reduce the negative effect of defective data in a driving behavior segmentation task. In the experiments, we used sticky hierarchical Dirichlet process hidden Markov model to segment the driving behavior. We compared the segmentation results using hidden features extracted by DSAE-FP and other comparative methods. Experimental results showed that segmentation results of non-defective dataset and defective dataset turned out most similar when DSAE-FP was used.","PeriodicalId":416104,"journal":{"name":"2016 IEEE 5th Global Conference on Consumer Electronics","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129928053","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 : 2016-10-01DOI: 10.1109/GCCE.2016.7800330
Susumu Genma, Takahiro Ogawa, M. Haseyama
This paper presents an image retrieval method for insect identification based on saliency map and distance metric learning. First, the proposed method extracts regions of insects from target images by using saliency map and calculates visual features from the extracted insect regions. Next, in order to realize accurate retrieval of insects based on the calculated features, distance metric learning is newly adopted. Consequently, through users' evaluation in the retrieval, optimal distance can be obtained for the calculated visual features to obtain successful retrieval results, and the identification of insects becomes feasible. Experimental results show the effectiveness of our method.
{"title":"Image retrieval for identification of insects based on saliency map and distance metric learning","authors":"Susumu Genma, Takahiro Ogawa, M. Haseyama","doi":"10.1109/GCCE.2016.7800330","DOIUrl":"https://doi.org/10.1109/GCCE.2016.7800330","url":null,"abstract":"This paper presents an image retrieval method for insect identification based on saliency map and distance metric learning. First, the proposed method extracts regions of insects from target images by using saliency map and calculates visual features from the extracted insect regions. Next, in order to realize accurate retrieval of insects based on the calculated features, distance metric learning is newly adopted. Consequently, through users' evaluation in the retrieval, optimal distance can be obtained for the calculated visual features to obtain successful retrieval results, and the identification of insects becomes feasible. Experimental results show the effectiveness of our method.","PeriodicalId":416104,"journal":{"name":"2016 IEEE 5th Global Conference on Consumer Electronics","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126476882","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 : 2016-10-01DOI: 10.1109/GCCE.2016.7800547
K. Chen, Yu-Shu Ni, Jhao-Yi Wang
Electrocardiography (ECG) is a widely used noninvasive clinical tool for the diagnosis of cardiovascular disease. However, the accuracy of ECG analysis significantly affect the diagnostic error rate of cardiovascular diseases. Therefore, in recent year, many Neural Network (NN)-based approaches were proposed to automatically analyze the ECG signal. However, these methods suffer from long computing time, which is inappropriate for the mobile real-time application. To solve this problem, we propose a Wavelet-based Artificial Neural Network (W-ANN) diagnosis flow in this paper. Based on the wavelet transform, the W-ANN can provide not only cleaner ECG input signal but lower computing time. The experimental results show that the proposed method can reduce 49% computing time with only 11.7% ECG diagnostic accuracy loss by involving the data from MIT-BIH arrhythmia database and real ECG signal measurement.
{"title":"Electrocardiogram diagnosis using wavelet-based artificial neural network","authors":"K. Chen, Yu-Shu Ni, Jhao-Yi Wang","doi":"10.1109/GCCE.2016.7800547","DOIUrl":"https://doi.org/10.1109/GCCE.2016.7800547","url":null,"abstract":"Electrocardiography (ECG) is a widely used noninvasive clinical tool for the diagnosis of cardiovascular disease. However, the accuracy of ECG analysis significantly affect the diagnostic error rate of cardiovascular diseases. Therefore, in recent year, many Neural Network (NN)-based approaches were proposed to automatically analyze the ECG signal. However, these methods suffer from long computing time, which is inappropriate for the mobile real-time application. To solve this problem, we propose a Wavelet-based Artificial Neural Network (W-ANN) diagnosis flow in this paper. Based on the wavelet transform, the W-ANN can provide not only cleaner ECG input signal but lower computing time. The experimental results show that the proposed method can reduce 49% computing time with only 11.7% ECG diagnostic accuracy loss by involving the data from MIT-BIH arrhythmia database and real ECG signal measurement.","PeriodicalId":416104,"journal":{"name":"2016 IEEE 5th Global Conference on Consumer Electronics","volume":"193 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122504380","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 : 2016-10-01DOI: 10.1109/GCCE.2016.7800311
Xin-Yu Shih, H. Chou, Yue-Qu Liu
FFT is a crucial and necessary design entity lor any current communication applications, especially for OFDM systems. As compared with conventional design approach, which can support only one-radix processing kernel of single-path delay feedback (SDF) FFT, we propose a reconfigurable processing kernel design for multiple-radix types of SDF FFT. In implementation, this kernel hardware engine is reconfigurable to deal with different computing stages and radix-types, such as 3-stage radix-2, 2-stage radix-3, and 1-stage radix-5 SDF FFT systems. Under TSMC 90nm CMOS technology, the kernel design with input/output wordlength of 16 bits only occupies 0.3068 mm2, saving 33.99% of area with respect to the summation of 3 individual single modes. Also, it only provides 13% hardware overhead as compared to pure radix-5 only processing kernel.
{"title":"Reconfigurable VLSI design of processing kernel for multiple-radix single-path delay feedback FFT systems","authors":"Xin-Yu Shih, H. Chou, Yue-Qu Liu","doi":"10.1109/GCCE.2016.7800311","DOIUrl":"https://doi.org/10.1109/GCCE.2016.7800311","url":null,"abstract":"FFT is a crucial and necessary design entity lor any current communication applications, especially for OFDM systems. As compared with conventional design approach, which can support only one-radix processing kernel of single-path delay feedback (SDF) FFT, we propose a reconfigurable processing kernel design for multiple-radix types of SDF FFT. In implementation, this kernel hardware engine is reconfigurable to deal with different computing stages and radix-types, such as 3-stage radix-2, 2-stage radix-3, and 1-stage radix-5 SDF FFT systems. Under TSMC 90nm CMOS technology, the kernel design with input/output wordlength of 16 bits only occupies 0.3068 mm2, saving 33.99% of area with respect to the summation of 3 individual single modes. Also, it only provides 13% hardware overhead as compared to pure radix-5 only processing kernel.","PeriodicalId":416104,"journal":{"name":"2016 IEEE 5th Global Conference on Consumer Electronics","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131247286","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 : 2016-10-01DOI: 10.1109/GCCE.2016.7800313
Chao-Yang Chang, Chung-Hsun Huang, Hui-Fu Chen, C. Yeh, Y. Chu, Tay-Jyi Lin
Overdrive technique is mandatory for liquid crystal display (LCD) to mitigate the motion blur phenomenon. As the display resolution increases, the image data should be highly compressed to reduce the usages of frame memory and bandwidth. Since current high quality compression algorithms inevitably require large embedded memory (line buffer) and complex computations, a low complexity line-buffer-free edge-preserved compression algorithm is proposed in this paper to reserve the edge information for overdrive technique of LCD. Performance evaluations show that the proposed edge-preserved compression achieves above 20% PSNR improvements as compared to the conventional DPCM compression up to 6:1 compression ratio.
{"title":"A low complexity edge-preserved image compression algorithm for LCD overdrive","authors":"Chao-Yang Chang, Chung-Hsun Huang, Hui-Fu Chen, C. Yeh, Y. Chu, Tay-Jyi Lin","doi":"10.1109/GCCE.2016.7800313","DOIUrl":"https://doi.org/10.1109/GCCE.2016.7800313","url":null,"abstract":"Overdrive technique is mandatory for liquid crystal display (LCD) to mitigate the motion blur phenomenon. As the display resolution increases, the image data should be highly compressed to reduce the usages of frame memory and bandwidth. Since current high quality compression algorithms inevitably require large embedded memory (line buffer) and complex computations, a low complexity line-buffer-free edge-preserved compression algorithm is proposed in this paper to reserve the edge information for overdrive technique of LCD. Performance evaluations show that the proposed edge-preserved compression achieves above 20% PSNR improvements as compared to the conventional DPCM compression up to 6:1 compression ratio.","PeriodicalId":416104,"journal":{"name":"2016 IEEE 5th Global Conference on Consumer Electronics","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131379348","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 : 2016-10-01DOI: 10.1109/GCCE.2016.7800482
Masayuki Kaneko, Kazuki Arima, Takashi Murakami, M. Isshiki, Hiroshi Sugimura
In this paper, we develop a system to superpose information of cooking onto kitchen counter. The system is controlled by natural actions which is voice, and is assembled by two major modules which a projector and the voice recognition. We describe the whole process, the prototype system and the evaluation.
{"title":"Development of superposing projection system for cooking support","authors":"Masayuki Kaneko, Kazuki Arima, Takashi Murakami, M. Isshiki, Hiroshi Sugimura","doi":"10.1109/GCCE.2016.7800482","DOIUrl":"https://doi.org/10.1109/GCCE.2016.7800482","url":null,"abstract":"In this paper, we develop a system to superpose information of cooking onto kitchen counter. The system is controlled by natural actions which is voice, and is assembled by two major modules which a projector and the voice recognition. We describe the whole process, the prototype system and the evaluation.","PeriodicalId":416104,"journal":{"name":"2016 IEEE 5th Global Conference on Consumer Electronics","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123948590","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 : 2016-10-01DOI: 10.1109/GCCE.2016.7800412
Ryota Nishimura, Yutaka Takase, Y. Nakano
In this paper, we describe the development of a spoken dialogue system based on PRINTEPS architecture. This spoken dialogue system is composed of five modules (speech recognition, language understanding, dialogue management, response generation, speech synthesis). In PRINTEPS, when calling the spoken dialogue system, system developers specify the small-scale dialogue goal. The system performs a dialogue with the user in order to obtain necessary information from the user. Dialogue processing rules corresponding to the dialogue goal is prepared in advance. The advantage of PRINTEPS-based system development is that system developers can build a spoken dialogue system without the knowledge of spoken dialogue system.
{"title":"Development environment of a spoken dialogue system based on PRINTEPS","authors":"Ryota Nishimura, Yutaka Takase, Y. Nakano","doi":"10.1109/GCCE.2016.7800412","DOIUrl":"https://doi.org/10.1109/GCCE.2016.7800412","url":null,"abstract":"In this paper, we describe the development of a spoken dialogue system based on PRINTEPS architecture. This spoken dialogue system is composed of five modules (speech recognition, language understanding, dialogue management, response generation, speech synthesis). In PRINTEPS, when calling the spoken dialogue system, system developers specify the small-scale dialogue goal. The system performs a dialogue with the user in order to obtain necessary information from the user. Dialogue processing rules corresponding to the dialogue goal is prepared in advance. The advantage of PRINTEPS-based system development is that system developers can build a spoken dialogue system without the knowledge of spoken dialogue system.","PeriodicalId":416104,"journal":{"name":"2016 IEEE 5th Global Conference on Consumer Electronics","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124041148","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 : 2016-10-01DOI: 10.1109/GCCE.2016.7800519
K. Kimura, Mikio Mori
In this study, we measured the thresholds of bone-conduction hearing via teeth while wearing a mouthpiece with a bone-conductive actuator and ear plugs. It was found that the thresholds of bone-conduction hearing via teeth tend to be better than those of hearing with no ear plugs.
{"title":"Effects of wearing ear plugs in bone-conduction hearing via teeth","authors":"K. Kimura, Mikio Mori","doi":"10.1109/GCCE.2016.7800519","DOIUrl":"https://doi.org/10.1109/GCCE.2016.7800519","url":null,"abstract":"In this study, we measured the thresholds of bone-conduction hearing via teeth while wearing a mouthpiece with a bone-conductive actuator and ear plugs. It was found that the thresholds of bone-conduction hearing via teeth tend to be better than those of hearing with no ear plugs.","PeriodicalId":416104,"journal":{"name":"2016 IEEE 5th Global Conference on Consumer Electronics","volume":"57 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127966400","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 : 2016-10-01DOI: 10.1109/GCCE.2016.7800531
T. Kajiyama
The image on a digital book cover gives potential buyers not only an impression of the book's contents but also serves as an aid when searching and browsing before and after buying the book. We have developed a method for revising book cover images to reflect readers' impressions. It uses a database of the correspondence between adjectives and colors as extracted from book reviews. This correspondence is used to revise the colors in the original book cover image. Evaluation was done by having 20 participants read 5 target books and give their impressions of the book cover images. The results revealed that the revised images better reflected the reader impressions than the original images.
{"title":"Color conversion of book cover images to reflect reader impressions","authors":"T. Kajiyama","doi":"10.1109/GCCE.2016.7800531","DOIUrl":"https://doi.org/10.1109/GCCE.2016.7800531","url":null,"abstract":"The image on a digital book cover gives potential buyers not only an impression of the book's contents but also serves as an aid when searching and browsing before and after buying the book. We have developed a method for revising book cover images to reflect readers' impressions. It uses a database of the correspondence between adjectives and colors as extracted from book reviews. This correspondence is used to revise the colors in the original book cover image. Evaluation was done by having 20 participants read 5 target books and give their impressions of the book cover images. The results revealed that the revised images better reflected the reader impressions than the original images.","PeriodicalId":416104,"journal":{"name":"2016 IEEE 5th Global Conference on Consumer Electronics","volume":"49 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115816150","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 : 2016-10-01DOI: 10.1109/GCCE.2016.7800541
Yang Yu, Z. Jia, Weige Tao, Shiliang Dong
Differential encoded LDPC (DE-LDPC) coded systems with multiple symbol differential detection (MSDD) scheme can avoid good estimation of channel state information and compensate performance degradation of differential detection. In this paper, LDPC codes optimization for the DE-LDPC coded systems with MSDD is studied. Simulation results show that the optimized LDPC codes perform much better than the regular LDPC codes for the DE-LDPC coded systems with MSDD.
{"title":"LDPC codes optimization for differential encoded LDPC coded systems with multiple symbol differential detection","authors":"Yang Yu, Z. Jia, Weige Tao, Shiliang Dong","doi":"10.1109/GCCE.2016.7800541","DOIUrl":"https://doi.org/10.1109/GCCE.2016.7800541","url":null,"abstract":"Differential encoded LDPC (DE-LDPC) coded systems with multiple symbol differential detection (MSDD) scheme can avoid good estimation of channel state information and compensate performance degradation of differential detection. In this paper, LDPC codes optimization for the DE-LDPC coded systems with MSDD is studied. Simulation results show that the optimized LDPC codes perform much better than the regular LDPC codes for the DE-LDPC coded systems with MSDD.","PeriodicalId":416104,"journal":{"name":"2016 IEEE 5th Global Conference on Consumer Electronics","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116897458","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：