Show and Tell Demonstrations
Location: Annex Hall
Wednesday, March 28, 10:30 – 12:30
ST-1.1 SIMPLE APPROACH TO ASSESSING PENILE ENDOTHELIAL FUNCTION IN YOUNG INDIVIDUALS
Hsien-Tsai Wu, Chun-Ho Lee, Jui-Ting Hsu, Chin-Jung Chen,and Cheuk-Kwan Sun
National Dong Hwa University, Hualien Hospital
Summary: Simple approach to assessing penile endothelial function in young individuals : Based on the physiological phenomenon of reactive hyperemia, systemic and penile vascular endothelial functions were assessed using digitally acquired data in this study. The results demonstrated highly significant correlation between the two functions in healthy young adults with normal penile erectile function. Application of APSS, therefore, provides a simple, economical, and convenient approach to the assessment of the risk for ED and cardiovascular diseases in young males. Briefly, the APSS system consists of two sets of pressure cuffs. The set over the upper arm is for triggering endothelial function, whereas the second set over the wrist is for data acquisition. A piezoresistive sensor (FGN-607PGSR, Fujilcura, Japan), which is connected to the second set of pressure cuff, is for pulse wave detection and arterial waveform recorded in the system. The system also contains an endothelial function measurement module board for obtaining, amplifying, and filtering the captured arterial waveform. Our results validated APSS as a reliable, economical, and easy-to-operate means of assessing penile endothelial function in young healthy adults for early identification of those at risk for ED.
ST-1.2 MULTISCALE ENTROPY ANALYSIS OF PULSE WAVE VELOCITY FOR ASSESSING ATHEROSCLEROSIS
Po-Chun Hsu, Kai-Yu Chang, Hou-Jun Wang, Hsien-Tsai Wu
National Dong Hwa University
Summary: lthough different means of non-invasive arterial stiffness assessment have been proposed, none is able to provide simultaneous information on whole-body peripheral arterial condition. The six-channel ECG-PWV measurement system showed remarkable applications in hypertensive subjects and with key anthropometric and biochemical parameters in diabetic patients, suggesting its validity in assessing arterial stiffness in subjects with peripheral arterial diseases. Moreover, given its capability of acquiring simultaneous PWVs from the body as a whole within 10 minutes and the ease of operation, it may serve as a convenient tool for early detection and following the progression of atherosclerosis in the population at risk. To achieve the same purpose of simultaneous whole-body assessment, we have previously proposed a six-channel ECG-PWV measurement instrument that can complete the calculations of six PWV values obtained from bilateral earlobes, fingers, and toes.
Wednesday, March 28, 14:00 – 16:00
ST-2.1 COMPRESSED SENSING PROTOTYPE FOR FLEXIBLE WIRELESS SYSTEM
Doohwan Lee, Takayuki Yamada, Kazunori Akabane, Yo Yamaguchi, Kazuhiro Uehara
NTT Corporation
Summary: This demonstration will show the implementation of flexible wireless system using compressed sensing (CS). FWS is a unified wireless platform which simultaneously deals with various types of wireless signals. Various types of wireless signals are received at a distributed access point (AP) and transferred to a server through the wired access line. Transferred signals are separated and demodulated at the server. To realize a highly flexible and efficient radio wave data reception and transmission between AP and the server, we applied compressed sensing. Furthermore, we also applied structured CS and 1-bit CS to reduce the processing burden and to enhance the compression performance. This demonstration will show the implementation of FWS under the framework of compressed sensing.
ST-2.3 CONFIGURABLE 3D AUDIO AND MUSIC
Qi (Peter) Li, Yin Ding, Manli Zhu
Li Creative Technologies, Inc.
Summary: An audience configurable 3D audio and music system is presented. Recently, we collected and processed a bank of highly accurate 3D head related transfer functions (HRTF). Given multiple channels of recorded music, the system can synthesis 3D music using the HRTF. In the demo, a person wearing a regular headset can feel that he or she stands in the center of a studio with four music instruments in the corners. A signer is singing a song while walking around him. This is a new way to enjoy music and no other system can provide such kind of performance and experience.
Thursday, March 29, 10:30 – 12:30
ST-3.1 REAL-TIME NOISE REDUCTION FOR DUAL-MICROPHONE MOBILE PHONES
Marco Jeub, Christian Herglotz, Hauke Krüger, Christophe Beaugeant, Peter Vary
RWTH Aachen University, Intel Mobile Communications France
Summary: We present a real-time demonstrator of noise reduction algorithms for dual-microphone mobile phones. Besides state-of-the-art algorithms, in particular a novel approach for dual-channel noise reduction is demonstrated which exploits the power level differences of speech and noise between the microphones. In the demonstrator, noisy speech signals which were recorded with a dummy head and dual-microphone mock-up phone as well as signals which are captured live during the demo session will be enhanced. Selected algorithm processing parameters can be modified during runtime to demonstrate the benefit of the new algorithm under realistic conditions.
ST-3.2 REAL-TIME AUDIO-VISUAL MEETING RECOGNITION AND UNDERSTANDING USING DISTANT MICROPHONE ARRAY
Takaaki Hori, Keisuke Kinoshita, Shoko Araki, Atsunori Ogawa, Takuya Yoshioka, Masakiyo Fujimoto, Takanobu Oba, Marc Delcroix, Mehrez Souden, Yotaro Kubo, Seong-Jun Hahm, Dan Mikami, Kazuhiro Otsuka, Tomohiro Nakatani, Atsushi Nakamura, Junji Yamato
NTT Corporation
Summary: We introduce an audio-visual meeting recognition system for monitoring conversations in an ongoing group meeting (http://www.kecl.ntt.co.jp/icl/signal/en/MeetingRecog.html). It automatically extracts such information as "who is speaking what to whom and when" in an online manner, by using state-of-the-art audio signal processing techniques including speech enhancement techniques (de-noising, BSS, dereverberation), ASR techniques (model adaptation, WSFT-based decoder) and image processing techniques proposed all by ourselves. The extracted information is displayed on the easily comprehensible browser implemented on a tablet-type smartphone. The presentation should be appealing to the ICASSP audience as the employed techniques covers a wide range of session topics in the conference.
ST-3.3 A MICROPHONE ARRAY FOR TABLET COMPUTERS AND SMART PHONES
Manli Zhu, Qi (Peter) Li
Li Creative Technologies, Inc.
Summary: A microphone array solution is presented to improve the speech communication for tablet computers, smart phones and other handheld videoconferencing devices. When a person uses a handheld device for videoconferencing, the voice quality in terms of SNR and intelligibility is significantly reduced as compared to a close-talking microphone. In order to solve this problem, we invented a microphone array system using adaptive beam-forming on 2-4 microphone components and the unique noise reduction algorithm. Our demo shows that both SNR and speech recognition accuracy can be significantly improved by using the array with real-time processing.
Thursday, March 29, 14:00 – 16:00
ST-4.1 HAPTIC VOICE RECOGNITION: A ROBUST MULTIMODAL INTERFACE FOR MOBILE DEVICES
Khe Chai SIM
National University of Singapore
Summary: Haptic Voice Recognition (HVR) is a novel multimodal text entry interface designed for modern portable devices equipped with microphone and touchscreen display. This interface enables users to enter text verbally, at the same time provide the initial letters of the spoken words using touch inputs to enhance the accuracy, efficiency and robustness. This proposed "Show & Tell" will demonstrate an HVR prototype implemented on the iOS platform. This prototype offers users two ways of entering the letter sequence: either by tapping on a soft keyboard or using handwriting gestures. More details about the HVR prototype can be found on http://speech.ddns.comp.nus.edu.sg/hvr/.
ST-4.2 HANNA'S TOURIST GUIDE FOR KYOTO
Teruhisa Misu, Etsuo Mizukami, Chiori Hori, Kiyonori Ohtake, Hideki Kashioka
National Institute of Information and Communications Technology (NICT)
Summary: We will show our spoken dialogue system that can explain tourist information of Kyoto. The system adopt statistical SLU module, which was realized using our language portability technique presented as a technical paper. The system is also expected to be downloaded from the apple store.
ST-4.3 NETWORK-BASED SPEECH-TO-SPEECH TRANSLATION FOR MULTIPARTY CHAT
Chiori Hori, Hideki Kashioka, Hisashi Kawai, Eichiro Sumita
National Institute of Information and Communications Technology (NICT)
Summary: To break language boundaries, Speech-to-Speech Translation technologies which translate speech in a source language to speech in a target language, is an effective means of communication between people who speak different languages. We introduce an application, the so-called ChaTra, which enables users to enjoy multilingual chats within the travel domain. The number of participate in a chat is confined from 2 to 5. The participants are able to select their favorable language from a list of 21 languages, including Japanese, English, Chinese, or Korean. Refer to more information at http://mastar.jp/chatra/chatra-e.html.