• Participation restrictions

  • You can participate as a team of multiple people.

  • Be sure to prepare the abbreviation name of each participating team. The abbreviation name is limited to 8 alphanumerical characters without spaces.

• What is lent to the participating teams at the registration

  • A program that recognizes the customer's position, voice, facial expression, gender, age, etc. using sensors.

  • A program for controlling Android I, a program for synthesizing the voice of Android I.

  • A simulator program for checking the behavior of Android I with CG during development.

  • Documents for using the above program.

  • A sample program using the above program.

  • Tourist information about 12 tourist destination candidates (using るるぶDATA * 1).

  • ※For some programs, the license key must be obtained and installed by the participating teams themselves (no cost).

• What the participating teams prepare for development

  • 1 microphone, 1 USB camera, PC for executing the above programs (1 Windows PC, 1 Linux PC with GPU).
    Recommended specifications for Linux PC

    • • OS: Ubuntu 16.04, 18.04

      • GPU: NVIDIA, memory 2GB or more

      • Main memory: 4GB or more

      • CPU: 8 cores or more

Recommended specifications for Windows PC

• • • • OS: Windows 10 64bit

      • Main memory: 8GB or more (16GB is enough)

      • CPU: 4 core Core i5 or better performance

  • Information on the customer's position, voice, facial expression, line of sight, body movement, gender / age (recognized based on the face image) recognized using the camera and microphone installed around Android I.

  • Information on the position of the monitor (where it is placed).

• What the participating teams do

  • Participating teams will develop a program that outputs the movements and utterances of Android I by inputting sensor information. You can use any language for development (Python, Java, C, etc.).

  • The recognition program using the sensor may be developed separately by the participating teams.

  • Participating teams can remotely test the developed programs using Android I before the preliminary round.

  • In the preliminary and final rounds, the participating teams are requested to bring their own PC (PC that runs the program developed by themselves). The sensors, Android I, and the PC that controls them will be the ones prepared at the venue.

  • Participating teams who have developed their own sensor-based recognition programs should also bring in the PCs needed to run them.

  • At the final round, each participating team will make presentations on the strategy of the system, dialogue generation algorithms, and ingenuity regarding the developed program. The teams that participated in the final round will make an oral presentation at the venue. Teams that only participate in the preliminary round will give a poster presentation at the venue.

• What the participating teams submit

  • The team description (2 pages, English) of the participating teams will be published as Proceedings. Participating teams should submit a team description summarized on page 2 regarding development policy for the developed program, dialogue generation algorithms, and ingenuity (the deadline will be after the final round).

  • Teams presenting at the final round should prepare the presentation slides in PowerPoint, PDF, etc. In addition, the presentation slides will be published on this website for the purpose of introducing the technologies created in this competition. The other teams should prepare a poster.