Top of page
Content Area
  • NTT Press Releases>
  • 2016>
  • NTT showcases its cutting-edge AI/IoT/video technologies at SXSW2016, the world’s largest music business and content festival

NTT Press Releases

March 3, 2016

Nippon Telegraph and Telephone Corporation

NTT showcases its cutting-edge AI/IoT/video technologies at SXSW2016, the world’s largest music business and content festival

SXSW (South by Southwest) 2016, the world’s largest business and content festival focusing on music, film and interactive topics will be held in Austin, Texas, USA. NTT Laboratories will present its “AI Showcase,” which will present its cutting-edge AI-related technologies and robots, and its “Interactive Live Showcase,” which uses IoT/video technologies, at JAPAN HOUSE (venue: The Vulcan Gas Company). This official event will be held on March 14 and 15, 2016 (local time). NTT will use this opportunity to demonstrate how valuable AI/IoT/video technologies can be in the robot/entertainment field. Going forward, NTT will proceed with commercialization by holding feasibility tests in which these technologies will be actually used in services and live events.


Overview of the exhibits showcased by NTT

(1)AI Showcase

Visitors can experience for themselves natural conversation with a computer by talking to an android (Geminoid HI-4) developed by Prof. Hiroshi Ishiguro, a world-renowned robotics expert. The android is powered by a variety of AI technologies developed by NTT Laboratories (clear sound pick-up in noisy environments, speech recognition, and casual conversation technologies). Even in the very noisy exhibition hall, visitors can effortlessly enjoy a conversation with CommU, a group of robots capable of conversation. This has been made possible by combining Osaka University’s technology, which makes robots interact with each other and respond to people, with NTT’s speech recognition and clear sound pickup technologies in noisy environments.

Prof. Hiroshi Ishiguro (left) and Geminoid HI-4 (right)Prof. Hiroshi Ishiguro (left) and Geminoid HI-4 (right)

Conversational robot group, CommUConversational robot group, CommU

[Exhibit information]

Visitors can see the above at the JAPAN HOUSE permanent exhibits “Autonomous Android: Geminoid HI-4” and “Conversational Robot Group Capable of Highly Natural Conversation: CommU.”
Also, at the stage session “Conversational Robot Technologies by Osaka Univ. and NTT featuring Dr. Ishiguro” (17:00-18:00 on March 14 and 15:00-16:00 on March 15), there will be a technical discussion session involving researchers from Prof. Ishiguro Laboratory who are at the forefront of AI development and researchers from NTT working on speech recognition technology.

Prof. Ishiguro will give a lecture entitled “Androids and Future Life” at the Austin Convention Center (12:30-13:30 on March 13). The lecture will include a demonstration of conversation in Japanese and English and conversation involving Japanese-English translation with an android (Geminoid HI-4) incorporating speech recognition, casual conversation, and machine translation technologies.

(2)Interactive Showcase

This exhibit will enable visitors to enjoy a novel live performance by Yun*chi, an ASOBISYSTEM artist and new-generation pop icon with a cute look and a clear voice. She has held live concerts in many countries. Her performance will be integrated with an entirely new sensation created using 2D pulse wave sensing technology for measuring pulse rate and technology of realistic floating images from a smartphone (Kirari! for Mobile), both developed by NTT Laboratories.
Specifically, the pulse rate data obtained in real time using a 2D pulse wave sensor will be reflected interactively on the background video on the live stage, creating a novel live performance world in which the artist and the audience will feel a sense of unity.
Paper craftwork made using the Kirari! for Mobile technology will be given away to the audience attending Yun*chi’s live performance. By attaching a craftwork to a smartphone, they can view vivid premium video content of Yun*chi for a limited time. They can have a glimpse of the future of entertainment supported by technology by watching a 3D video of Yun*chi on their hands and basking in the afterglow of her performance.


[Exhibit information]

Visitors can experience these technologies at Stage Session “Yun*chi Interactive LIVE Showcase Presented by NTT” (21:00-22:00 on March 15).

The features of the technologies showcased in these exhibits are described in the attachment.

Note:Overview of JAPAN HOUSE
Theme : Extension of Humanity
Dates : Monday, March 14 - Tuesday, March 15, 2016
Time : 9:00 - 26:00 on either day
Venue : The Vulcan Gas Company (Austin, Texas, USA)
418 E. 6th Street, Austin, TX 78701

(Attachment) Features of the technologies

(1)AI-related technologies (clear sound pick-up in noisy environments, speech recognition, casual conversation, and machine translation)

- Clear sound pick-up technology in noisy environments
This technology uses intelligent microphone technology developed by NTT Laboratories. Clear sound pick-up in any noisy environments is realized by combining a small number of microphones and sound signal processing. The latter uses the directions which sounds are coming from, frequency properties and temporal variations in these properties to reduce the magnitude of ambient noise while minimizing degradation in the target sound, thereby making it possible to pick up the target sound clearly. This technology allows highly accurate speech recognition even in noisy environments.
- Speech recognition technology
Even if the target speech involves the use of a large vocabulary, this technology achieves high-speed, high-accuracy speech recognition by using acoustic modeling based on the latest deep learning and an optimal search network based on the weighted finite state transducer (WFST).
- Casual conversation technology
This technology makes conversation over a wide range of topic possible by using language processing to structuralize a large volume of text data and by building up a reservoir of casual conversation “knowledge” from the data. Naturalness of dialog is strengthened by casual conversation breakdown technology, which accurately suppresses any speech that is deemed inappropriate for the current context.
- Machine translation technology
Highly accurate machine translation is realized by using three technologies: language analysis technology that trains a highly precise and compact model from a large volume of text data, advance sorting technology that resolves differences in word orders in different languages, and statistical machine translation technology that acquires translation skills automatically from a large volume of bilingual data.

(2)IoT-related technology (2D pulse wave sensing technology)

This technology makes it possible to measure a pulse wave (variations in the blood flow volume in a peripheral artery) just by holding a mouse or the steering wheel of a vehicle without having to wear any kind of dedicated device. It uses an optical principle to achieve a wider sensing range than conventional sensors, allowing users to measure their pulse rate while going about their usual daily activities without having to pay attention to the position of the sensor.

2D pulse wave sensor-mounted microphone2D pulse wave sensor-mounted microphone

(3)Video-related technology: Kirari! for Mobile (Technology of realistic floating images from a smartphone)

Kirari! for Mobile is a device that enables users to haverealistic floating images viewing experience that makes them feel as if they were present at the scene displayed just by attaching a smartphone. Even though the device is small, a highly realistic space representation is possible because videos can be projected on multiple layers with natural depth intervals between them.

  • Appearance of Kirari! for MobileAppearance of Kirari! for Mobile
  • Viewing using Kirari! for MobileViewing using Kirari! for Mobile

Contact Information

Research and Development Planning Department,R&D planning Group

Nippon Telegraph and Telephone Corporation

Innovative R&D by NTT
NTT Has Instituted a Logo to Represent R&D Activities.

Information is current as of the date of issue of the individual press release.
Please be advised that information may be outdated after that point.

NTT Press Releases Index