-
原稿種別: 付録等
p.
App1-
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
-
本山 史明, 東海 彰吾, 森岡 貴義
原稿種別: 本文
p.
1-1-1_-_1-1-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
Abstract We consider a method for collecting of orientation information of smartphones of multi-view shooting system to analyze situations of the shooting scene in real time. In this consideration, collective-observed areas are obtained based on overlapping situation of smartphones' viewing fields which are collected through a network of the system.
抄録全体を表示
-
今西 健児, 伊藤 義道, 中村 和晃, 小野 士, 馬場口 登
原稿種別: 本文
p.
1-2-1_-_1-2-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
We propose a method for estimating user location histories in an indoor environment. To estimate the user location, we use sensors in a smartphone and fixed camera images.
抄録全体を表示
-
清水 智行
原稿種別: 本文
p.
1-3-1_-_1-3-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
HTTP streaming typically requires media stream division into segments. Shorter segment could reduce latency, while it might produce header overhead and cause degradation of TCP throughput estimation. Thus we investigated influence of media stream division on HTTP streaming by analyzing media segment structure and actual throughput evaluation.
抄録全体を表示
-
原田 臨太朗, 黒川 沙代, 竹中 幸子, 武藤 健史, 金井 謙治, 甲藤 二郎
原稿種別: 本文
p.
1-4-1_-_1-4-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
We evaluated various characteristics of 4K video streaming using DASH-JS over LTE, public Wi-Fi services and WiMAX. We did experiments while stopping, walking, and riding the train. We found that a mobile user could enjoy smooth video streaming by using these networks even 4K video.
抄録全体を表示
-
金子 豊, 竹内 真也, 黄 民錫, 苗村 昌秀
原稿種別: 本文
p.
1-5-1_-_1-5-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
We developed a novel zapping system with tag for viewing TV program in the time shifting environment. This paper describes the overview of this prototype system.
抄録全体を表示
-
柴田 賀昭
原稿種別: 本文
p.
1-6-1_-_1-6-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
A project is ongoing at SMPTE to enhance applications of UMID. While the latest SMPTE RP 205 specifying the UMID Application Principles will be published shortly, the UMID Resolution Protocol is under standardization. In this paper, we demonstrate how UMIDs in a MXF file will be exploited to realize various applications based on those standards.
抄録全体を表示
-
山崎 俊彦, 前田 嵩樹, 相澤 清晴
原稿種別: 本文
p.
2-1-1_-_2-1-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
We have been proposing a two-stage classification framework for image recognition that conjunctively uses parametric and non-parametric approaches, which we call inclined local naive Bayes nearest neighbor (NBNN). In this paper, our method is incooperated into state-of-the-art bag of features (BoF) based approaches and demonstrate that the proposed method is general and can outperform most of BoF-based methods.
抄録全体を表示
-
高松 幸広, 相澤 清晴, 小川 誠
原稿種別: 本文
p.
2-2-1_-_2-2-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
We have released a smartphone application, in iOS and Android, named FoodLog for personal food recording purposes. In FoodLog data, a distribution of food item positions in an image was observed. In this paper, we make use of the food positions to improve the accuracy of food image retrieval.
抄録全体を表示
-
CHOI Saemi, YAMASAKI Toshihiko, AIZAWA Kiyoharu
原稿種別: 本文
p.
2-3-1_-_2-3-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
We introduce 20 questions game into a personal image retrieval system. Our goal is to retrieve user intended photograph and to minimize the number of questions by not asking unnecessary questions which are discriminated using context data such as time-stamps and geolocation of the personal photograph database.
抄録全体を表示
-
Du Bang, 近藤 一晃, 吉本 廣雅, 中村 裕一
原稿種別: 本文
p.
2-4-1_-_2-4-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
In working situations, we are often sensing and interacting with more than one target simultaneously. It means that our attention is distributed to different targets in different ways. For analyzing such diversity, we had preliminary experiments for measuring not only eye-gazing but also a head, torso, and other body portions and movements in typical cooking work.
抄録全体を表示
-
河合 吉彦, 望月 貴裕, 住吉 英樹, 藤原 忍
原稿種別: 本文
p.
2-5-1_-_2-5-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
This paper proposes a video retrieval method based on object recognition. The object recognition is a technique of identifying shown object in a video frame using machine learning and it enables user to search desired scenes considering the video content. Experimental results showed the effectiveness of the proposed method.
抄録全体を表示
-
望月 貴裕, 河合 吉彦, 住吉 英樹
原稿種別: 本文
p.
2-6-1_-_2-6-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
When the East Japan great earthquake disaster occurred about three years ago, a very large amount, over 20,000 hours, of video news material was recorded in a short time. We broadcaster have important missions to make effective use of this video for the disaster prevention and mitigation in the future. Currently, the trial introduction of an Earthquake Disaster Archives Retrieval System is under way, which is based on automatic metadata extraction by our image analysis technology. Additionally, the function to retrieve video-shots based on the visual-based image similarity is scheduled to be installed. In this paper, we describe the brief introduction of current Earthquake Disaster Archives Retrieval System and the detail of visual-based similar image retrieval technology as a new function.
抄録全体を表示
-
SABIRIN Houari, SANKOH Hiroshi, NAITO Sei
原稿種別: 本文
p.
3-1-1_-_3-1-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
In this paper we propose a method of automatic object tracking for soccer players based on the best match of objects' attributes, such as location, dominant color and motion vector, to accurately detect and track soccer players especially in the event of object occlusions. Each detected object in the soccer video is assigned attributes that represent the information extracted from the pixel values of the segmented object. The correlations of their attributes are then observed to find the similarity between objects in two consecutive frames. By observing the best match on attribute similarity, the proposed method can reliably track the objects and consistently provide the object with unique identification even during the event of occlusion as shown in the results of the experiment
抄録全体を表示
-
百合本 瑞規, 榎並 直子, 有木 康雄
原稿種別: 本文
p.
3-2-1_-_3-2-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
In this paper, we propose a method to estimate a flow of each lane and vehicle position in the lane from images by car mounted stereo camera for active navigation. The effectiveness is evaluated with the Karlsruhe dataset of images captured by the car-mounted stereo cameras in real environments.
抄録全体を表示
-
黒川 陽平, 中島 悠太, 佐藤 智和, 横矢 直和
原稿種別: 本文
p.
3-3-1_-_3-3-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
In this paper, toward light-weight camera pose estimation for mobile devices, we propose a new error criterion for camera poses without similarity measure-based matching between detected feature points and 3D points in point cloud database.
抄録全体を表示
-
佐藤 恩子, 田川 憲男, 大久保 寛, 何 宜欣
原稿種別: 本文
p.
3-4-1_-_3-4-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
We have proposed a 3D recovery method using image blur caused by camera rotations imitating involuntary eye movements. Image blur appears high-frequency components prominently. Therefore, elimination of low-frequency components is expected to improve the sensitivity of image blur detection. In this study, we evaluate such the effect through numerical simulations.
抄録全体を表示
-
石川 諭, 小林 祐一, 金子 透, 石原 進, 山下 淳
原稿種別: 本文
p.
3-5-1_-_3-5-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
This paper presents a study on generation of an image using two images taken at shifted locations by in-vehicle cameras. Generated image is expected to help intuitive understanding of traffic situations for drivers through the use of vehicular ad hoc network.
抄録全体を表示
-
土出 衛, 仲野 豊, 吉田 俊之
原稿種別: 本文
p.
3-6-1_-_3-6-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
Shape From Focus, which has been known as a depth estimation technique based on focus information, often yields very poor depth estimation in highlighting and/or very dark regions. This paper proposes a technique that overcomes the problem by using multi-exposure image sets. The experimental results show that the proposed technique is promising.
抄録全体を表示
-
赤井 亮太, 新田 直子, 馬場口 登
原稿種別: 本文
p.
3-7-1_-_3-7-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
People counting in a wide area requires establishing correspondence among human groups traveling across non-overlapping camera views. Iteratively establishing weighted many-to-many correspondence among the human groups traveling within a short time interval realizes the real-time and accurate people counting even when human groups split or merge outside the camera views.
抄録全体を表示
-
安藤 俊博, 亀田 裕介, 松田 一朗, 伊東 晋
原稿種別: 本文
p.
3-8-1_-_3-8-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
For the purpose of user interface or marketing, estimation of gaze directions by a noncontact measurement device has been studied in recent years. This paper describes an eigenspace-based method which estimates horizontal and vertical angles of the gaze directionfrom eye region images.
抄録全体を表示
-
中村 有揮, 松田 一朗, 亀田 裕介, 伊東 晋
原稿種別: 本文
p.
3-9-1_-_3-9-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
We previously proposed a semi-automatic 2D-3D conversion method for still images. The method propagates depth information, which is partially given by the user as an input, to a whole image based on local similarity of image values. Therefore, the propagation process must be sufficiently fast for interactive operation. In this paper, we accelerate this process by using a kind of distance transform.
抄録全体を表示
-
川原 功, 田端 宏至
原稿種別: 本文
p.
4-1-1_-_4-1-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
A quantitative yet visual assessment of moving image quality for 4K8K UHDTV systems including HEVC CODECs is developed. Using dedicated patterns under various motion, efficiency and reliability of the evaluation are secured by taking full advantage of human potential in perception, while highly considering display ergonomics, or visual comfort.
抄録全体を表示
-
窪田 悟
原稿種別: 本文
p.
4-2-1_-_4-2-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
We investigated the influence of the angular size of display on the preferred display luminance and viewing gamma. The results indicate that the preferred luminance decreases and the preferred viewing gamma increases as the display viewing angle increases. These results can be applied to provide an optimum picture quality for large 4K displays.
抄録全体を表示
-
日下部 裕一, 正岡 顕一郎, 西田 幸博
原稿種別: 本文
p.
4-3-1_-_4-3-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
We have studied the viewing conditions for subjective assessment of picture quality of "Super Hi-Vision (SHV)", which features high resolution and wide field of view. We propose that three evaluators are aligned at a viewing distance of 0.75 times the display height and positioned at the lateral trisection points of the display width to evaluate SHV pictures. We conducted a subjective evaluation of SHV picture quality and observed a significant difference in the mean opinion scores at the different viewing positions, which demonstrates the validity of our proposal.
抄録全体を表示
-
和泉 紘介, 吉野 知伸, 内藤 整, 高本 考一, 西条 寿夫
原稿種別: 本文
p.
4-4-1_-_4-4-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
Quality of Experience(QoE)is important for video service using high quality video content. To evaluate QoE based on questionnaire is difficult because evaluation score varies due to personal equation. To overcome this problem, some researches assessed the quality based on biomedical responses. In this paper, we focused on measuring brain hemodynamic activity using a NIRS (Near-infrared spectroscopy) method. We analyzed relationships between the brain activity and video resolutions.
抄録全体を表示
-
井上 哲理, 柴田 隆史
原稿種別: 本文
p.
4-5-1_-_4-5-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
We examined a sense of height in virtual environment by using 3D CG models of cityscape displayed on a CAVElike display system or a wide-view type HMD device. The experiments were conducted to evaluate effects of density and height of surrounding models of building on the subjects' sense of height.
抄録全体を表示
-
加藤 幸洋, 柿本 剛志, 井上 康之, 河合 敦夫, 井須 尚紀
原稿種別: 本文
p.
4-6-1_-_4-6-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
Biological signals were analyzed with the aim of predicting the onset of cyber sickness before VE users feel discomfort. Subjects operated a driving simulator for 40 minutes to induce cyber sickness. They evaluated time-course variations of discomfort, enjoyment and relaxation. Covariance structure of biological signals and subjective evaluations was analyzed.
抄録全体を表示
-
石先 広海, 服部 元, 滝嶋 康弘
原稿種別: 本文
p.
5-1-1_-_5-1-1_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
In this paper, we propose a background music player based on scene understanding with images taken by smart glass, in order to enrich user experience with appropriate background music. We demonstrate our system which enables user to listen to background music, recommended based on real-time scene understanding on a smart grass.
抄録全体を表示
-
川原 功, 田端 宏至
原稿種別: 本文
p.
5-2-1_-_5-2-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
A quantitative yet visual assessment of moving image quality for 4K8K UHDTV systems including HEVC CODECs is developed. Using dedicated patterns under various motion, efficiency and reliability of the evaluation are secured by taking full advantage of human potential in perception, while highly considering display ergonomics, or visual comfort.
抄録全体を表示
-
河邉 隆寛, 澤山 正貴, 丸谷 和史, 西田 眞也
原稿種別: 本文
p.
5-3-1_-_5-3-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
We report a light projection method to illusorily deform static objects in printed materials by motion information that is projected through a video projector. By employing this method, it is possible to add the material impression of non-rigid materials to the printed materials.
抄録全体を表示
-
小田 周平, 遠藤 洋介
原稿種別: 本文
p.
6-1-1_-_6-1-1_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
We propose a Diffserv-based video network where Quality of Service(QoS) is managed with differentiated priority. To operate the QoS, we develop a network management and monitoring system that allows active rate control for users.
抄録全体を表示
-
田中 壮, 西村 敏, 遠藤 洋介
原稿種別: 本文
p.
6-2-1_-_6-2-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
We have proposed a push content delivery method of broadcast program related data aimed at simultaneous reception in each client. This method is effective under the condition that clients' RTT deviation of a delivery server is high. In this report, we propose a server selection method which assigns clients to each server by considering the RTT distribution of each server.
抄録全体を表示
-
成清 善一, 宮坂 宏明, 中村 円香, 実井 仁, 高田 政幸
原稿種別: 本文
p.
6-3-1_-_6-3-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
We are developing a next-generation terrestrial broadcasting transmission system for handheld and mobile reception which uses two transmission antennas and two receiving antennas (Multiple-Input Multiple-Output: MIMO) and can transmit large capacity contents stably. This paper shows the field experiment results of the SFBC-MIMO and the SDM-MIMO system. As the result, the SDM-MIMO system reception performance is better than SFBC-MIMO system when the transmission rate is 4bits/Hz and more.
抄録全体を表示
-
土生谷 隆弘, 早見 拓朗, 藤田 欣裕
原稿種別: 本文
p.
6-4-1_-_6-4-1_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
Using TVML (TV-program Making Language), an experimental service system for TV and tablet PC s has been developed. Also, a practical and simple synchronizing method for broadcast data and those from communication channel has been proposed. We have examined that even with a small data channel bandwidth, this system showed good performance and very effective.
抄録全体を表示
-
堀 明宏
原稿種別: 本文
p.
7-1-1_-_7-1-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
Conventional video system is based on the peak luminance limitation of CRT, 〜100cd/m2.It is possible to provide the luminance beyond this limitation with LCD. Video dynamic range preference was surveyed by the viewer's preference. The result shows the peak luminance 10,000cd/m2 is more preferable. High Dynamic Range video system provides better quality video.
抄録全体を表示
-
神谷 雅志, 中村 雄大, 福田 智教, 内藤 正博, 南 浩次
原稿種別: 本文
p.
7-2-1_-_7-2-1_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
We evaluated usability of controlling a device with hand gesture recognition. We found that an acceptable response time is less than 110 ms, and 250 ms are limited.
抄録全体を表示
-
原田 直人, 張 英夏, 向井 信彦
原稿種別: 本文
p.
7-3-1_-_7-3-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
In this paper, we propose a method to recognize Japanese finger characters on hand images by using a decision tree. The recognition rate in an experiment for 41 Japanese finger characters without movement was 96.5%.
抄録全体を表示
-
小野 士, 中村 和晃, 馬場口 登
原稿種別: 本文
p.
7-4-1_-_7-4-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
In this paper, we propose Information Entry System (IES) for personalized services in real environments. IES is a system for collecting users' privacy information by image analysis and providing the users with an opportunity to control the disclosure level of their privacy information through a touch-screen monitor.
抄録全体を表示
-
貴志 一樹, 山崎 俊彦, 相澤 清晴
原稿種別: 本文
p.
7-5-1_-_7-5-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
Lifelog, logging the whole data of daily life is getting popular recently. In this paper, we propose a method to recognize the action of deskwork which has relation to many people by using skeleton and image features. The manually labeling cost reduction method is also discussed. By labeling 1% of whole frames, about 90% labels are correctly estimated.
抄録全体を表示
-
山内 結子, 奥田 誠, 高橋 正樹, サイモン クリピングデル, 苗村 昌秀
原稿種別: 本文
p.
7-6-1_-_7-6-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
We are researching technologies for acquisition and analysis of viewing status and the estimation of viewers' interests in TV viewing environment, with a view to developing new services based on viewers' preferences. We report results from a TV viewing experiment conducted in a living-room environment, and describe a prototype system for estimation of the viewers' interest based on attentional behavior.
抄録全体を表示
-
奥田 誠, 高橋 正樹, クリピングデル サイモン, 山内 結子, 苗村 昌秀
原稿種別: 本文
p.
7-7-1_-_7-7-1_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
We aim to develop a method of inferring words that represent the interests of TV viewers. We analyzed the relationship among TV viewers' facial expression, closed captions, and interest words. The results showed a tendency for interest words to be present in closed captions displayed in scenes when the viewers' faces were expressive.
抄録全体を表示
-
高山 伸也, 酒澤 茂之, 愛澤 伯友
原稿種別: 本文
p.
7-8-1_-_7-8-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
This paper presents the evaluation results of education contents using 3-D Character. Experimental results show that the contents with captions are more effective for improving student comprehension than those without captions.
抄録全体を表示
-
YU Siyang, KONDO Kazuaki, YOSHIMOTO Hiromasa, NAKAMURA Yuichi, DANTSUJ ...
原稿種別: 本文
p.
7-9-1_-_7-9-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
In e-learning, learner's attitude, i.e. how learner learns is useful information for teachers to improve the course. We designed a system that captures learner's behaviors to estimate learner's concentration status, and a user interface for easy browse of concentration information.
抄録全体を表示
-
野上 隆司, 片野 正徳, 小池 中
原稿種別: 本文
p.
8-1-1_-_8-1-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
This video recording system automatically predicts the beginning of an earthquake according to Earthquake Early Waning and takes out the moment quickly from its server.
抄録全体を表示
-
佐藤 庄衛, 尾上 和穂, 小林 彰夫, 奥 貴裕, 藤田 悠哉, 一木 麻乃, 荒井 孝
原稿種別: 本文
p.
8-2-1_-_8-2-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
We propose an automatic captioning method. The Method compares the program manuscript with the automatically recognized speech and instantly identifies the line in the manuscript corresponding to the speech. This makes it possible to eliminate manual error correction in the production of closed-captions for broadcasting.
抄録全体を表示
-
宮崎 正之, 河野 勝, 杉浦 智保, 大長 久芳
原稿種別: 本文
p.
8-3-1_-_8-3-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
White-LEDs with Cl_MS phosphor can reduce the unpleasant glare and improve the color renderings compared with the conventional white-LEDs. Then, we developed studio illuminators using white-LEDs with Cl_MS phosphor. In this paper, we introduce the evaluation results about these illuminators and the trial results in the TV program production spots.
抄録全体を表示
-
栗山 和久
原稿種別: 本文
p.
8-4-1_-_8-4-1_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
This system makes it possible to switch between two cameras with suitable background image.
抄録全体を表示
-
栗山 和久
原稿種別: 本文
p.
8-5-1_-_8-5-1_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
This vehicle makes it possible to transmit video quickly using satellite and WiFi.
抄録全体を表示
-
浅野 隼, 日下部 武志, 小田 周平
原稿種別: 本文
p.
8-6-1_-_8-6-2_
発行日: 2014/09/01
公開日: 2017/05/24
会議録・要旨集
フリー
We developed a remote storage transmission system that automatically gathers high-quality video signals at the moment of an earthquake in best-effort IP network. We evaluated the transfer rate of the video signals and the required time from the trigger of earthquake information to the transmission completion of the video signals in actual FTTH network.
抄録全体を表示