Couger develops Human-like AI assistant that communicates like a human. Testing its potential as a next-generation interface on digital signage around train stations.

Couger Inc. (Headquarters: Shibuya-ku, Tokyo; CEO: Atsushi Ishii) , in collaboration with Information Services Int'l-Dentsu, Ltd. (Headquarters: Minato-ku, Tokyo; President: Ryoichi Nawa; hereinafter "ISID"), will conduct a demonstration test of Couger's human-like AI assistant "VHA (Virtual Human Agent)" at the commercial facility "Trier Keio Chofu'' located near Chofu Station on the Keio Line on June 29 and 30, 2019.

In this experiment, a life-size Human-like AI assistant will be projected on an outdoor digital signage screen, and will communicate with passersby. The interaction will include not only verbal communication but also non-verbal communication such as emotional expressions and hand gestures, which people do unconsciously, to see how it affects their behavior.

Keio Corporation (Head office: Tama City, Tokyo; President: Yasushi Benimura), the operator of Triennial Chofu, the AI assistant will communicate the attractiveness of Chofu and information about the surrounding area, and will verify whether it can contribute to the improvement of local love and revitalization of the city.

Couger expects that VHA to produce results in a variety of areas in the future, including advertising, entertainment, and tourism.

<VHA interacts with people through facial expressions and movements>

■Background: Redefining the Customer Experience with Human-like AI Assistants

VHA enhances the customer experience by becoming an agent that connects companies and customers.

Currently, the means by which stores and companies provide information to customers include one-way transmissions via TV, magazines, etc., and the transmission of recommendations and advertisements based on user behavior on the Internet.

The problem with these methods of providing information is that there is overwhelmingly little "dialogue" between the two parties, making it extremely difficult to provide appropriate information to each customer.

In a human dialogue, the person in question may smile and say, "You look happy," or express concern while asking, "Are you feeling alright?” In every situation, humans communicate by obtaining information such as emotions from the other person's face and body movements, thereby creating empathy and deep communication.

The combination of these emotions, facial expressions, gestures, and other verbal and non-verbal parts of speech provide an important part of approaching human interaction. VHA achieves this by utilizing real-time video recognition technology, natural language processing, and other technologies.

As automation and unmanned operations continue to advance in all industries around the world, VHA is committed to being humanoid, expressing a human-like sense of security and familiarity, and providing customer-oriented communication. In the U.S., several studies have been conducted through humanoid interfaces, and research has shown (*1) that humanoid interfaces displaying faces and bodies have improved both the speed and content of human responses compared to voice-only interfaces.

Through VHA, we will go back to the basics of human communication and aim to reinvent the customer experience based on the original human communication.

■Outline and points of the demonstration experiment

<Outline of implementation>

Date : June 29 (Sat.) and 30 (Sun.), 2019, 10:00 a.m. to 6:00 p.m.

Place : In front of Building C, Trier Keio Chofu (2-61-1 Kojimacho, Chofu City, Tokyo)

How to participate :No advance registration is required and everyone is welcome to attend. (Language: Japanese only)

<Experiment flow and key points>

  • The persona of the VHA appearing on the digital signage is "a chatty older sister who knows Chofu well.”
  • When the person is standing by, the VHA uses attractive actions and changes in facial expression to draw attention to the person.
  • Once the person is recognized, wave to him/her or greet him/her to start a conversation.
  • Conversation with participants through active speech, such as self-introductions
  • Lead a dialogue by quizzing participants on Chofu's attractions and events, and informing them of local attractions and activities
  • Finally, ask participants to take action.

<Verification items>

  • Whether or not VHA can create communication
  • Analysis of communication content
  • Change in user behavior through communication
  • Evaluation of social acceptability, convenience and impression of VHA

※1 References

1. "The Effect of An Animated Virtual Character on Mobile Chat Interactions" 2

2. "The LISSA Virtual Human and ASD Teens: An Overview of Initial Experiments"

<For more information, please contact us.>

Inquiries via Web

Tel: 03-6433-5226Attn: Tatsumi

Recommended articles