Ritch 2.0 vs PhyOSC

Ritch is an interactive narrative and questionnaire application, guided by an Avatar assistant. The avatar and tasks inside the application help reinforce coping skills to severe issues like drug abuse and lack of anger management. The application allows users to create an account which sends that data up a secure database, stored with no identifying information of the user. The application then uses that account to save question and quiz data to report the findings back to our researchers.

PhyOSC is an application that supports your exercise with Apple Watch, controlling the sound according to your body movement. Instead of uniform music at the sports gym, The instructor's improvisational movement enhances the atmosphere of the venue during the session. ------------------------- Features of PhyOSC ------------------------- 1) PhyOSC receives Core Motion data from Apple Watch on iPhone. The iPhone converts this sensor data and 12 types of arm movements trained by machine learning into OSC messages and sends them to the IP address and Port number to which the DAW is connected. 2) PhyOSC can apply a digital filter when converting Apple Watch Core Motion data received by iPhone into OSC message data. 3) OSC can name and save the filtered settings. This setting can be instantly recalled and switched, so even with the same arm movement, setting A produces a cheering sound effect, and switching to setting B produces a drum kick sound. It realizes a variety of expressions such as changing the waveform of the oscillator and the oscillator. In addition, the TCP / IP address and Port number of the OSC message destination are set in the settings, so you can switch DAW devices by switching the settings. ------------------------- System Requirements ------------------------- PhyOSC requires iPhone and Apple Watch. The required operating environment is as follows: Apple Watch : watchOS 8 Apple Watch Series 5 or higher recommended iPhone: iOS 15 iPhone 8 or higher recommended ------------------------- Limitations ------------------------- The accuracy of machine learning recognition is not 100%. Incorrect results may be output depending on the speed and angle of arm movement. We continue to research the latest research and look for new approaches to improve accuracy.