发布求助
文献互助 智能选刊 最新文献

2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR)最新文献

筛选
英文 中文
Deep learning using heterogeneous feature maps for maxout networks 基于异构特征映射的maxout网络深度学习
2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR) Pub Date : 2015-11-01 DOI: 10.1109/ACPR.2015.7486545
Yasunori Ishii, Reiko Hagawa, Sotaro Tsukizawa
{"title":"Deep learning using heterogeneous feature maps for maxout networks","authors":"Yasunori Ishii, Reiko Hagawa, Sotaro Tsukizawa","doi":"10.1109/ACPR.2015.7486545","DOIUrl":"https://doi.org/10.1109/ACPR.2015.7486545","url":null,"abstract":"We propose a novel type of maxout that uses filters with kernels of multiple sizes for generating convolved maps. These filters extract the most effective features for recognition from many different variations of texture patterns. A convolved map is generated by convolution between feature maps and filters. If the size of filters is varied, the size of the convolved map will also vary; in which case, since there are no correspondences among the positions of convolved maps, maxout will not work for these kinds of filters. To align the sizes of convolved maps, we converted, in advance, feature maps, which we term `heterogeneous feature maps,' using zero padding. Converting the size of feature maps in this way allows maxout to function, even with filters of different sizes. In this study we demonstrate the classification performances using our proposed maxout on MNIST, CIFAR-10, CIFAR-100, SVHN datasets, and show a 13.17% improvement of accuracy with augmented data.","PeriodicalId":240902,"journal":{"name":"2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR)","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128467793","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}
引用次数: 1
Unsupervised daily routine modelling from a depth sensor using top-down and bottom-up hierarchies 无监督的日常建模从深度传感器使用自上而下和自下而上的层次结构
2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR) Pub Date : 2015-11-01 DOI: 10.1109/ACPR.2015.7486465
Yangdi Xu, David Bull, D. Damen
{"title":"Unsupervised daily routine modelling from a depth sensor using top-down and bottom-up hierarchies","authors":"Yangdi Xu, David Bull, D. Damen","doi":"10.1109/ACPR.2015.7486465","DOIUrl":"https://doi.org/10.1109/ACPR.2015.7486465","url":null,"abstract":"A person's routine incorporates the frequent and regular behaviour patterns over a time scale, e.g. daily routine. In this work we present a method for unsupervised discovery of a single person's daily routine within an indoor environment using a static depth sensor. Routine is modelled using top down and bottom up hierarchies, formed from location and silhouette spatio-temporal information. We employ and evaluate stay point estimation and time envelopes for better routine modelling. The method is tested for three individuals modelling their natural activity in an office kitchen. Results demonstrate the ability to automatically discover unlabelled routine patterns related to daily activities as well as discard infrequent events.","PeriodicalId":240902,"journal":{"name":"2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124776936","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}
引用次数: 2
An improved segmentation of online English handwritten text using recurrent neural networks 一种改进的基于递归神经网络的在线英语手写文本分割方法
2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR) Pub Date : 2015-11-01 DOI: 10.1109/ACPR.2015.7486489
C. Nguyen, M. Nakagawa
{"title":"An improved segmentation of online English handwritten text using recurrent neural networks","authors":"C. Nguyen, M. Nakagawa","doi":"10.1109/ACPR.2015.7486489","DOIUrl":"https://doi.org/10.1109/ACPR.2015.7486489","url":null,"abstract":"Segmentation of online handwritten text recognition is better to employ the dependency on context of strokes written before and after it. This paper shows an application of Bidirectional Long Short-term Memory recurrent neural networks for segmentation of on-line handwritten English text. The networks allow incorporating long-range context from both forward and backward directions to improve the confident of segmentation over uncertainty. We show that applying the method in the semi-incremental recognition of online handwritten English text reduces up to 62% of waiting time, 50% of processing time. Moreover, recognition rate of the system also improves remarkably by 3 points from 71.7%.","PeriodicalId":240902,"journal":{"name":"2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR)","volume":"155 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126844861","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}
引用次数: 8
Hough-based action detection with time-warped voting 基于时间扭曲投票的hough动作检测
2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR) Pub Date : 2015-11-01 DOI: 10.1109/ACPR.2015.7486581
Kensho Hara, K. Mase
{"title":"Hough-based action detection with time-warped voting","authors":"Kensho Hara, K. Mase","doi":"10.1109/ACPR.2015.7486581","DOIUrl":"https://doi.org/10.1109/ACPR.2015.7486581","url":null,"abstract":"Hough-based action detection methods cast weighted votes for action classes and positions based on the local spatio-temporal features of the given video sequences. Conventional Hough-based methods perform poorly for actions with temporal variations because such variations change the temporal relation between the local feature positions and the global action positions. Some votes may scatter because of such variations. In this paper, we propose a method for concentrating scattered votes through a time warping of the votes. The proposed method calculates the offsets between the scattered voting positions and the concentrated positions based on the votes generated through the conventional Hough-based method. The offsets warp the scattered votes to concentrate them, and provide a method of robustness even in the presence of temporal variations. We experimentally confirmed that the proposed method improves the average precision for the UT-Interaction dataset compared with a conventional method.","PeriodicalId":240902,"journal":{"name":"2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR)","volume":"142 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123461602","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}
引用次数: 0
Feature extraction with convolutional restricted boltzmann machine for audio classification 基于卷积受限玻尔兹曼机的音频分类特征提取
2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR) Pub Date : 2015-11-01 DOI: 10.1109/ACPR.2015.7486611
Min Li, Z. Miao, Cong Ma
{"title":"Feature extraction with convolutional restricted boltzmann machine for audio classification","authors":"Min Li, Z. Miao, Cong Ma","doi":"10.1109/ACPR.2015.7486611","DOIUrl":"https://doi.org/10.1109/ACPR.2015.7486611","url":null,"abstract":"Feature extraction is a crucial part for a large number of audio tasks. Researchers have extracted audio features in multiple ways, among which some most recent methods are based on the hidden layer of a trained neutral network. In this paper, we present a system which can automatically extract features from unlabeled audio data, and then the features of extracted from the system are used for audio classification task. Ourfeature extraction scheme makes use of a convolutional restricted Boltzmann machine (CRBM), instead of those using restricted Boltzmann machines (RB-M). By using features extracted from CRBM, we can achieve about 7% accuracy improvement consistently over than the RBM-based features on the TI-Digits dataset for audio classification. We also combine the well-known MFCC features and the CRBM-based features in the form of a linear combination. In our experiments, this feature combining the two methods performs better than both features alone.","PeriodicalId":240902,"journal":{"name":"2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR)","volume":"158 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132336013","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}
引用次数: 7
Trajectory-based stereo visual odometry with statistical outlier rejection 具有统计离群值抑制的基于轨迹的立体视觉里程计
2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR) Pub Date : 2015-11-01 DOI: 10.1109/ACPR.2015.7486575
Jiyuan Zhang, Rui Gan, Gang Zeng, Falong Shen, H. Zha
{"title":"Trajectory-based stereo visual odometry with statistical outlier rejection","authors":"Jiyuan Zhang, Rui Gan, Gang Zeng, Falong Shen, H. Zha","doi":"10.1109/ACPR.2015.7486575","DOIUrl":"https://doi.org/10.1109/ACPR.2015.7486575","url":null,"abstract":"We present a stereo visual odometry algorithm with trajectorical information accumulated over time and consistency among multiple trajectories of different motions. The objective function considers transfer error of all previously observed points to reduce drifting, and can be efficiently approximated and optimized within a computational bound. Different from traditional residual-based consistency measurement, we exploit the linear system in non-linear optimization to evaluate the influence of each point for outlier rejection. Both the drifting and irruptive error are reduced by combining trajectorical information of multiple motions. Experiments with real world dataset show that our method could handle difficult scenes with large portion of outliers without expensive computation.","PeriodicalId":240902,"journal":{"name":"2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130006522","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}
引用次数: 0
Depth-based person re-identification 基于深度的人物再识别
2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR) Pub Date : 2015-11-01 DOI: 10.1109/ACPR.2015.7486459
Ancong Wu, Weishi Zheng, J. Lai
{"title":"Depth-based person re-identification","authors":"Ancong Wu, Weishi Zheng, J. Lai","doi":"10.1109/ACPR.2015.7486459","DOIUrl":"https://doi.org/10.1109/ACPR.2015.7486459","url":null,"abstract":"Person re-identification aims to match people across non-overlapping camera views. For this purpose, most works exploit appearance cues, assuming that the color of clothes is discriminative in short term. However, when people appear in extreme illumination or change clothes, appearance-based methods tend to fail. Fortunately, depth images provide more invariant body shape and skeleton information regardless of illumination and color, but only a few depth-based methods have been developed so far. In this paper, we propose a covariance-based rotation invariant 3D descriptor called Eigen-depth to describe pedestrian body shape and the property of rotation invariance is proven in theory. It is also insensitive to slight shape change and invariant to color change and background. We combine our descriptor with skeleton-based feature to get a complete representation of human body. The effectiveness is validated on RGBD-ID and BIWIRGBD-ID datasets.","PeriodicalId":240902,"journal":{"name":"2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR)","volume":"172 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132618223","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}
引用次数: 5
Linear multimodal fusion in video concept analysis based on node equilibrium model 基于节点平衡模型的视频概念分析中的线性多模态融合
2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR) Pub Date : 2015-11-01 DOI: 10.1109/ACPR.2015.7486517
Jie Geng, Z. Miao, Qinghua Liang, Shu Wang
{"title":"Linear multimodal fusion in video concept analysis based on node equilibrium model","authors":"Jie Geng, Z. Miao, Qinghua Liang, Shu Wang","doi":"10.1109/ACPR.2015.7486517","DOIUrl":"https://doi.org/10.1109/ACPR.2015.7486517","url":null,"abstract":"Multiple modalities such as color, texture, shape and motion need to be analyzed separately and fused together to get the comprehensive result in content-based video concept analysis. We propose a multimodal fusion method based on a mechanical node equilibrium model. It treats the scores ofmultiple modalities and the fused score as physical nodes. Between these nodes, we define correlations which are treated as forces to move the nodes to a new position. Finally, the whole node system will be at an equilibrium status which is regarded as the fusion result. Essentially, the proposed method is a linear fusion model with linear fusion equations. The correlations are optimized by an expectation maximum (EM) algorithm which is quite efficient needing only several iterations.","PeriodicalId":240902,"journal":{"name":"2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132807423","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}
引用次数: 0
Real-time fingertip detection based on depth data 基于深度数据的实时指尖检测
2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR) Pub Date : 2015-11-01 DOI: 10.1109/ACPR.2015.7486542
Chaoyu Liang, Yonghong Song, Yuanlin Zhang
{"title":"Real-time fingertip detection based on depth data","authors":"Chaoyu Liang, Yonghong Song, Yuanlin Zhang","doi":"10.1109/ACPR.2015.7486542","DOIUrl":"https://doi.org/10.1109/ACPR.2015.7486542","url":null,"abstract":"In this paper we propose a novel method to detect fingertip using depth data. The first step of our method is to segment hand from depth map precisely. Then a two layer hand model is constructed to detect self-occlusion and mitigate its impact. In the next step an extended graph model of hand is built to locate and label finger bases. Then we generate heat maps of finger bases to detect finger regions even fingers are closed or adhesion occurs. Finally fingertips are located on fingers by geodesic paths. Experiments on different finger motions and hand rotations show that our framework performs accurately when hand pose and rotation change. Compared with other approaches our method shows less errors and robust to depth noise.","PeriodicalId":240902,"journal":{"name":"2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR)","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131444292","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}
引用次数: 3
Neural network based over-segmentation for scene text recognition 基于神经网络的场景文本过分割识别
2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR) Pub Date : 2015-11-01 DOI: 10.1109/ACPR.2015.7486596
Xin He, Yi-Chao Wu, Kai Chen, Fei Yin, Cheng-Lin Liu
{"title":"Neural network based over-segmentation for scene text recognition","authors":"Xin He, Yi-Chao Wu, Kai Chen, Fei Yin, Cheng-Lin Liu","doi":"10.1109/ACPR.2015.7486596","DOIUrl":"https://doi.org/10.1109/ACPR.2015.7486596","url":null,"abstract":"Over-segmentation is often used in text recognition to generate candidate characters. In this paper, we propose a neural network-based over-segmentation method for cropped scene text recognition. On binarized text line image, a segmentation window slides over each connected component, and a neural network is used to classify whether the window locates a segmentation point or not. We evaluate several feature representations for window classification and combine sliding window-based segmentation with shape-based splitting. Experimental results on two benchmark datasets demonstrate the superiority and effectiveness of our method in respect of segmentation point detection and word recognition.","PeriodicalId":240902,"journal":{"name":"2015 3rd IAPR Asian Conference on Pattern Recognition (ACPR)","volume":"18 12","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131687435","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}
引用次数: 3
首页 上一页 下一页 尾页
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整，为了账户安全，请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
请完成安全验证×
相关产品
×
本文献相关产品
联系我们：info@booksci.cn Book学术提供免费学术资源搜索服务，方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号：604180095
Book学术官方微信