SiteSCAN
March 2021-September 2021, 6 months
Yu Chen Chen, NTU
Kai Hsin Chiu, NTU
Tsai Ping Chang, NTUST
Li Fei Kung, NTU
Role | Researcher
Tools | Figma
Teammates |
A system that reshapes the way people create and interact with the information through introducing an intuitive design solution to exchange and explore spatial information.
The main goal is to eliminate the alienation between spatial information online and the real physical world.
Paper accepted by ACM MobileHCI 2021 Demos and Interactivity.
Background
Nowadays, as the explosion of information creates a sense of alienation between people and the information, it is hard for us to link the information on the internet (virtual information) to real-world settings. And even when on-site, there’s barely a way to directly get information regarding the place; the existing forms of information fail to describe their relation to space precisely.
Overview
▲ How SiteSCAN works!
▲ Quick view of SiteSCAN!
Design Concept & Design Highlights
Use retrieval technology and geolocation positioning to identify location.
To create information:
To create some new information about the space, users can add information to the system by scanning the surroundings, creating information stickers, and then attaching them to specific spots in space.
To explore information:
Users can view the information by simply scanning the spot and the InfoSticker would render on where they were attached to, reviving their relation to the space with high fidelity.
Challenges: How does this community work?
This community mainly relies on crowdsourcing, in which user-contributed information will keep updating and growing gradually as users stay active and increase over time.
Results
The AR lenses we came up with no longer leave users alienated between virtual and physical reality and support users’ access to information as they move around without the hassle of having to switch cross-websites or even cross-platforms.
Moving forward, we need to introduce an alternative way to integrate information other than crowdsourcing, for its limitation depending on manually and on-site by users and better allow information to come from existing sources ( e.g., database, websites). Also as the community grows, we have to make a filter system introduced to filter and limit the search results shown to the users. Last but not least, we would like to enhance the image matching performance through a better deep learning model.