Pub Date : 2016-12-01DOI: 10.1109/APSIPA.2016.7820713
Fang-Yu Chao, Jia Xu, Chia-Wen Lin
In this paper, we propose a framework that fuses textual and visual features of user generated social media data to mine the distribution of user interests. The proposed framework consists of three steps: feature extraction, model training, and user interest mining. We choose boards from popular users on Pinterest to collect training and test data. For each pin we extract the term-document matrices as textual features, bag of visual words as low-level visual features, and attributes as mid-level visual features. Representative features are then selected for training topic models using discriminative latent Dirichlet allocation (DLDA). In performance evaluation, pins collected from popular users are used to evaluate the classification accuracy and pins collected from other common users are used to evaluate the recommendation performance. Our experimental results show the efficacy of the proposed method.
{"title":"Mining user interests from social media by fusing textual and visual features","authors":"Fang-Yu Chao, Jia Xu, Chia-Wen Lin","doi":"10.1109/APSIPA.2016.7820713","DOIUrl":"https://doi.org/10.1109/APSIPA.2016.7820713","url":null,"abstract":"In this paper, we propose a framework that fuses textual and visual features of user generated social media data to mine the distribution of user interests. The proposed framework consists of three steps: feature extraction, model training, and user interest mining. We choose boards from popular users on Pinterest to collect training and test data. For each pin we extract the term-document matrices as textual features, bag of visual words as low-level visual features, and attributes as mid-level visual features. Representative features are then selected for training topic models using discriminative latent Dirichlet allocation (DLDA). In performance evaluation, pins collected from popular users are used to evaluate the classification accuracy and pins collected from other common users are used to evaluate the recommendation performance. Our experimental results show the efficacy of the proposed method.","PeriodicalId":409448,"journal":{"name":"2016 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA)","volume":"56 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134038941","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 : 2016-12-01DOI: 10.1109/APSIPA.2016.7820774
Yunseok Song, Yo-Sung Ho
Time-of-Flight (ToF) cameras are easily accessible in this era. They capture real distances of objects in a controlled environment. Yet, the ToF image may include disconnected boundaries between objects. In addition, certain objects are not capable of reflecting the infrared ray such as black hair. Such problems are caused by the physics of ToF. This paper proposes a method to compensate such errors by replacing them with reasonable distance data. The proposed method employs object boundary filtering, outlier elimination and iterative min/max averaging. After acquiring the enhanced ToF image, this can be applied to depth map generation by using the ToF camera with other color cameras. The experiment results show improved ToF images which lead to more accurate depth maps.
{"title":"Time-of-flight image enhancement for depth map generation","authors":"Yunseok Song, Yo-Sung Ho","doi":"10.1109/APSIPA.2016.7820774","DOIUrl":"https://doi.org/10.1109/APSIPA.2016.7820774","url":null,"abstract":"Time-of-Flight (ToF) cameras are easily accessible in this era. They capture real distances of objects in a controlled environment. Yet, the ToF image may include disconnected boundaries between objects. In addition, certain objects are not capable of reflecting the infrared ray such as black hair. Such problems are caused by the physics of ToF. This paper proposes a method to compensate such errors by replacing them with reasonable distance data. The proposed method employs object boundary filtering, outlier elimination and iterative min/max averaging. After acquiring the enhanced ToF image, this can be applied to depth map generation by using the ToF camera with other color cameras. The experiment results show improved ToF images which lead to more accurate depth maps.","PeriodicalId":409448,"journal":{"name":"2016 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA)","volume":"147 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133364498","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 : 2016-12-01DOI: 10.1109/APSIPA.2016.7820679
A. Sasou
Voice-pathology detection from a subject's voice is a promising technology for pre-diagnosis of larynx diseases. Glottal source estimation in particular plays a very important role in voice-pathology analysis. For more accurate estimation of the spectral envelope and glottal source of the pathology voice, we propose a method that can automatically generate the topology of the glottal source Hidden Markov Model (HMM), as well as estimate the Auto-Regressive (AR)-HMM parameter by combining AR-HMM parameter estimation and the Minimum Description Length-based Successive State Splitting (MDL-SSS) algorithm. The AR-HMM adopts a single Gaussian distribution for the output Probability Distribution Function (PDF) of each state in the glottal source HMM. In this paper, we propose a novel voice-pathology detection method based on the AR-HMM with automatic topology generation, which utilizes the output PDF variances normalized with regard to the maximum variance as clues for voice-pathology detection. We experimentally demonstrate that for normal voices, other normalized variances are distributed around a lower range than the maximum variance. This is because the PDF of the state just following vocal fold closure tends to have a maximum variance far greater than other variances. For pathology voices, the maximum variance and other variances are more closely distributed than for normal voices, possibly due to air leaking through the vocal folds. The experiment results confirmed the feasibility and fundamental validity of the proposed method.
{"title":"Voice-pathology analysis based on AR-HMM","authors":"A. Sasou","doi":"10.1109/APSIPA.2016.7820679","DOIUrl":"https://doi.org/10.1109/APSIPA.2016.7820679","url":null,"abstract":"Voice-pathology detection from a subject's voice is a promising technology for pre-diagnosis of larynx diseases. Glottal source estimation in particular plays a very important role in voice-pathology analysis. For more accurate estimation of the spectral envelope and glottal source of the pathology voice, we propose a method that can automatically generate the topology of the glottal source Hidden Markov Model (HMM), as well as estimate the Auto-Regressive (AR)-HMM parameter by combining AR-HMM parameter estimation and the Minimum Description Length-based Successive State Splitting (MDL-SSS) algorithm. The AR-HMM adopts a single Gaussian distribution for the output Probability Distribution Function (PDF) of each state in the glottal source HMM. In this paper, we propose a novel voice-pathology detection method based on the AR-HMM with automatic topology generation, which utilizes the output PDF variances normalized with regard to the maximum variance as clues for voice-pathology detection. We experimentally demonstrate that for normal voices, other normalized variances are distributed around a lower range than the maximum variance. This is because the PDF of the state just following vocal fold closure tends to have a maximum variance far greater than other variances. For pathology voices, the maximum variance and other variances are more closely distributed than for normal voices, possibly due to air leaking through the vocal folds. The experiment results confirmed the feasibility and fundamental validity of the proposed method.","PeriodicalId":409448,"journal":{"name":"2016 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA)","volume":"150 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"113960604","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 : 2016-12-01DOI: 10.1109/APSIPA.2016.7820837
N. T. Khajavi, A. Kuh
We conduct a study of graphical models and discuss the quality of model selection approximation by formulating the problem as a detection problem and examine the area under the curve (AUC). We are specifically looking at the model selection problem for jointly Gaussian random vectors. For Gaussian distributions, this problem simplifies to the covariance selection problem which is widely discussed in literature by Dempster [1]. In this paper, we discuss graphical models such as the pth order Markov chain and the pth order star network interpretation which also have junction tree graphical representations and give the definition for the correlation approximation matrix (CAM) which contains all information about the model selection problem. We compute the model covariance matrix as well as the KL divergence between the original distribution and the approximated model distribution. We conduct some simulations which show that the quality of the selected model increases as the model order, p, increases.
{"title":"The covariance selection quality for graphs with junction trees through AUC bounds","authors":"N. T. Khajavi, A. Kuh","doi":"10.1109/APSIPA.2016.7820837","DOIUrl":"https://doi.org/10.1109/APSIPA.2016.7820837","url":null,"abstract":"We conduct a study of graphical models and discuss the quality of model selection approximation by formulating the problem as a detection problem and examine the area under the curve (AUC). We are specifically looking at the model selection problem for jointly Gaussian random vectors. For Gaussian distributions, this problem simplifies to the covariance selection problem which is widely discussed in literature by Dempster [1]. In this paper, we discuss graphical models such as the pth order Markov chain and the pth order star network interpretation which also have junction tree graphical representations and give the definition for the correlation approximation matrix (CAM) which contains all information about the model selection problem. We compute the model covariance matrix as well as the KL divergence between the original distribution and the approximated model distribution. We conduct some simulations which show that the quality of the selected model increases as the model order, p, increases.","PeriodicalId":409448,"journal":{"name":"2016 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA)","volume":"52 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122945613","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 : 2016-12-01DOI: 10.1109/APSIPA.2016.7820706
Liyuan Xiong, Wei Zhou, Xin Zhou, Guanwen Zhang, Ai Qing
In this paper, a saliency aware fast intra coding algorithm for HEVC is proposed consists of perceptual intra coding and fast intra prediction mode decision algorithm. Firstly, based on the visual saliency detection, an adaptive CU splitting method is proposed to reduce intra encoding complexity. Furthermore, quantization parameter is adaptively adjusted at the CU level according to the relative importance of each CU and distortion is efficiently controlled. Secondly, a fast intra prediction mode decision algorithm with step halving rough mode decision method and early modes pruning algorithm is presented to selectively check the potential modes and effectively reduce the complexity of computation. Experimental results show that 45.39% encoding time can be reduced by the proposed saliency aware fast intra coding algorithm. Furthermore, our proposed algorithm can achieves 2.18% bit rate reduction on average with negligible perceptual quality loss.
{"title":"Saliency aware fast intra coding algorithm for HEVC","authors":"Liyuan Xiong, Wei Zhou, Xin Zhou, Guanwen Zhang, Ai Qing","doi":"10.1109/APSIPA.2016.7820706","DOIUrl":"https://doi.org/10.1109/APSIPA.2016.7820706","url":null,"abstract":"In this paper, a saliency aware fast intra coding algorithm for HEVC is proposed consists of perceptual intra coding and fast intra prediction mode decision algorithm. Firstly, based on the visual saliency detection, an adaptive CU splitting method is proposed to reduce intra encoding complexity. Furthermore, quantization parameter is adaptively adjusted at the CU level according to the relative importance of each CU and distortion is efficiently controlled. Secondly, a fast intra prediction mode decision algorithm with step halving rough mode decision method and early modes pruning algorithm is presented to selectively check the potential modes and effectively reduce the complexity of computation. Experimental results show that 45.39% encoding time can be reduced by the proposed saliency aware fast intra coding algorithm. Furthermore, our proposed algorithm can achieves 2.18% bit rate reduction on average with negligible perceptual quality loss.","PeriodicalId":409448,"journal":{"name":"2016 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA)","volume":"08 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128990296","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 : 2016-12-01DOI: 10.1109/APSIPA.2016.7820726
Kenta Iida, H. Kiya
A secure identification scheme for JPEG 2000 code-streams is proposed in this paper. The aim is to securely identify JPEG 2000 images generated from the same original image, without decoding images. Features used for the identification are extracted from header parts in a JPEG 2000 codestream. The proposed scheme does not provide any false negative matches under various compression ratios, while most of image hashing-based schemes do not guarantee this performance. Existing identification schemes that do not provide any false negative matches can not be securely carried out. Due to such a situation, we propose an identification system based on a fuzzy commitment scheme, which is a well-known secure protocol for biometric template protection. Moreover, an error correction technique with 1-bit parity is considered to achieve the system. The experiment results show the proposed scheme is effective in terms of true positive matches, while keeping the security high.
{"title":"Codestream level secure identification for JPEG 2000 images under various compression ratios","authors":"Kenta Iida, H. Kiya","doi":"10.1109/APSIPA.2016.7820726","DOIUrl":"https://doi.org/10.1109/APSIPA.2016.7820726","url":null,"abstract":"A secure identification scheme for JPEG 2000 code-streams is proposed in this paper. The aim is to securely identify JPEG 2000 images generated from the same original image, without decoding images. Features used for the identification are extracted from header parts in a JPEG 2000 codestream. The proposed scheme does not provide any false negative matches under various compression ratios, while most of image hashing-based schemes do not guarantee this performance. Existing identification schemes that do not provide any false negative matches can not be securely carried out. Due to such a situation, we propose an identification system based on a fuzzy commitment scheme, which is a well-known secure protocol for biometric template protection. Moreover, an error correction technique with 1-bit parity is considered to achieve the system. The experiment results show the proposed scheme is effective in terms of true positive matches, while keeping the security high.","PeriodicalId":409448,"journal":{"name":"2016 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129313831","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 : 2016-12-01DOI: 10.1109/APSIPA.2016.7820897
Yong-Jun Chang, Yo-Sung Ho
Stereo matching methods estimate depth information of captured images. One way to estimate accurate depth values is to use the distance information. This method enhances the disparity map by preserving the edge region. In order to preserve the depth discontinuity near the edge region, it uses the distance information as a new weighting value for the matching cost function. However, this method has a high complexity problem. To overcome this problem, we propose region based stereo matching method with gradient and distance information. Since the distance transform calculates the pixel distance from the edge region, we can classify whether the pixel is near the edge region or not. In other words, some regions near the edge have small distance transformed values. For this reason, our method divides regions depending on the value of distance transformed pixel. After that, different cost functions are applied to each region for improving the computation efficiency.
{"title":"Region based stereo matching method with gradient and distance information","authors":"Yong-Jun Chang, Yo-Sung Ho","doi":"10.1109/APSIPA.2016.7820897","DOIUrl":"https://doi.org/10.1109/APSIPA.2016.7820897","url":null,"abstract":"Stereo matching methods estimate depth information of captured images. One way to estimate accurate depth values is to use the distance information. This method enhances the disparity map by preserving the edge region. In order to preserve the depth discontinuity near the edge region, it uses the distance information as a new weighting value for the matching cost function. However, this method has a high complexity problem. To overcome this problem, we propose region based stereo matching method with gradient and distance information. Since the distance transform calculates the pixel distance from the edge region, we can classify whether the pixel is near the edge region or not. In other words, some regions near the edge have small distance transformed values. For this reason, our method divides regions depending on the value of distance transformed pixel. After that, different cost functions are applied to each region for improving the computation efficiency.","PeriodicalId":409448,"journal":{"name":"2016 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA)","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116084946","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 : 2016-12-01DOI: 10.1109/APSIPA.2016.7820807
Hirotaka Mukumoto, K. Hayashi, Megumi Kaneko
Direction-of-arrival (DOA) estimation via Khatri-Rao (KR) subspace with multiple signal classification (MUSIC) algorithm can cope with a higher number of incoming waves than that of sensors, while it requires the signals to be quasi-stationary and needs a larger number of “frames” than that of incoming waves. On the other hand, a hybrid approach of compressed sensing and MUSIC algorithm can estimate DOAs with snapshots less than the number of sources for noiseless observation, although the number of incoming waves must be less than that of sensors. Exploiting the fact that the frame in MUSIC via KR subspace corresponds to the snapshot in conventional MUSIC without observation noise, we propose a DOA estimation scheme using KR product array processing and compressed sensing, which can cope with a greater number of incoming waves than both that of sensors and that of frames. The validity of the proposed method is shown via numerical experiments.
{"title":"Direction-of-arrival estimation via Khatri-Rao subspace using compressed sensing","authors":"Hirotaka Mukumoto, K. Hayashi, Megumi Kaneko","doi":"10.1109/APSIPA.2016.7820807","DOIUrl":"https://doi.org/10.1109/APSIPA.2016.7820807","url":null,"abstract":"Direction-of-arrival (DOA) estimation via Khatri-Rao (KR) subspace with multiple signal classification (MUSIC) algorithm can cope with a higher number of incoming waves than that of sensors, while it requires the signals to be quasi-stationary and needs a larger number of “frames” than that of incoming waves. On the other hand, a hybrid approach of compressed sensing and MUSIC algorithm can estimate DOAs with snapshots less than the number of sources for noiseless observation, although the number of incoming waves must be less than that of sensors. Exploiting the fact that the frame in MUSIC via KR subspace corresponds to the snapshot in conventional MUSIC without observation noise, we propose a DOA estimation scheme using KR product array processing and compressed sensing, which can cope with a greater number of incoming waves than both that of sensors and that of frames. The validity of the proposed method is shown via numerical experiments.","PeriodicalId":409448,"journal":{"name":"2016 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA)","volume":"55 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115625910","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 : 2016-12-01DOI: 10.1109/APSIPA.2016.7820772
Li-Ang Chen, Jian-Jiun Ding, Yih-Cherng Lee
As the annoying blocking or ghost artifacts tend to appear in the conventional compression approaches either in the JPEG or JPEG2000 standards at low bitrate, the concept of the object-oriented image compression is proposed. This kind of methods is able to retain the image structural boundaries and therefore has relatively good visual qualities even in high compression ratios. In this paper, we propose a shape-adaptive image compression scheme employing an efficient lossy contour compression algorithm to encode the region information, which is usually the main overhead data in such systems. In addition, the prediction and deblocking techniques commonly used in novel compression approaches are also applied with the proposed shape-adaptive versions. Simulation results suggest that the proposed compression system is able to provide compressed images with much better visual qualities and more reasonable degradation forms compared to other prevailing methods.
{"title":"Shape-adaptive image compression using lossy shape coding, SA-prediction, and SA-deblocking","authors":"Li-Ang Chen, Jian-Jiun Ding, Yih-Cherng Lee","doi":"10.1109/APSIPA.2016.7820772","DOIUrl":"https://doi.org/10.1109/APSIPA.2016.7820772","url":null,"abstract":"As the annoying blocking or ghost artifacts tend to appear in the conventional compression approaches either in the JPEG or JPEG2000 standards at low bitrate, the concept of the object-oriented image compression is proposed. This kind of methods is able to retain the image structural boundaries and therefore has relatively good visual qualities even in high compression ratios. In this paper, we propose a shape-adaptive image compression scheme employing an efficient lossy contour compression algorithm to encode the region information, which is usually the main overhead data in such systems. In addition, the prediction and deblocking techniques commonly used in novel compression approaches are also applied with the proposed shape-adaptive versions. Simulation results suggest that the proposed compression system is able to provide compressed images with much better visual qualities and more reasonable degradation forms compared to other prevailing methods.","PeriodicalId":409448,"journal":{"name":"2016 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114602740","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 : 2016-12-01DOI: 10.1109/APSIPA.2016.7820756
F. Asano, Miyuki Fukushima
The evacuation of children and the elderly from disaster areas is sometimes difficult. This study aims to use a vibration sensor to estimate situations involving people who remain in a devastated building. This paper proposes a method to estimate the attributes of the people, such as their age or sex, based on the vibration data produced by their footsteps. The vibration data obtained through sensors are analyzed by a linear prediction method to extract the features, which are then classified using a support vector machine to estimate the attributes. The experimental results show that an accuracy of 80–95% was achieved for the classification of the sex and the type of shoes.
{"title":"Classification of footstep attributes using a vibration sensor","authors":"F. Asano, Miyuki Fukushima","doi":"10.1109/APSIPA.2016.7820756","DOIUrl":"https://doi.org/10.1109/APSIPA.2016.7820756","url":null,"abstract":"The evacuation of children and the elderly from disaster areas is sometimes difficult. This study aims to use a vibration sensor to estimate situations involving people who remain in a devastated building. This paper proposes a method to estimate the attributes of the people, such as their age or sex, based on the vibration data produced by their footsteps. The vibration data obtained through sensors are analyzed by a linear prediction method to extract the features, which are then classified using a support vector machine to estimate the attributes. The experimental results show that an accuracy of 80–95% was achieved for the classification of the sex and the type of shoes.","PeriodicalId":409448,"journal":{"name":"2016 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA)","volume":"157 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115179515","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：