The project is KNUA(Korea National University of Arts) a fusion arts center, <team running 2019: Hey, Strangers. Do It Together!>.This is the output of team "Faint Hope”.
In this project, we defined expressions which have two contrasting emotions as ambiguous expressions of a lot of expression mixed with more than two emotions. Analyzing ambiguousness of emotions through API, we developed this project as a game to rethink about ambiguousness by comparison and crossing about recognizing process of other’s emotion in each perspective of human and AI.
Add a Teacher Note to share how you incorporated it into your lesson.
Step 1: NEEDS
1. processing download : https://processing.org
2. Microsoft Azure account : https://processing.org
Installing Processing on PC
1. Access the processing homepage and install processing appropriate to your operating system
1. click processing you downloaded. 2. Execute processing Processing IDE -> Sketch -> Internal Library ... -> Adding Library ... click. 3. When the Contribution Manager is launched, select the Libraries tab and search for HTTP Requests. 4. Select HTTP Requests for Processing library and click Install. 5. When installation is complete, quit processing and restart (recommended)
How to make Microsoft Azure account
1. connect to MS Azure https://azure.microsoft.com/
1. Go to portal after signing in to Azure. https://azure.microsoft.com/
3. To create a new resource, choose Create Resource -> AI + Machine Learning -> Face Resource Selection.
4. Click Start Trial for your trial account.
5. Enter the resource information and click Create.
step 3. Check the address and key of the API.
1. Click on the resource that you registered and get the key -> Click the key
2. Confirm your key.
3. Check the access address of Face API.
Step 2: [Reference] MS Azure Pricing Policy
Step 3: Let's Try It!
1. Download the File.
2. Enter your own key in the example.
3. click play button.
4. Click on the display window and wait.
5. check the result!!!!!!
Step 5: COMMENTS & REGRETS, We Need Help!!
In addition to images, we tried to analyze the ambiguous values derived from text, such as Twitter. We also wanted the format of the game, users were presented with a ranking method, so that they could continue to use the game and get them interested about ‘what makes ‘ ambiguos(we call EMEMOHO)''. Based on the relationship between expressions and emotions, we aimed to enable users to explore and define ambiguity.
cording skills. we did not reflect the research data for utilizing Twitter. Change image library construction method - build libraries by Google crawling in real-time, but it was replaced by people's labor because of lack of ability. Even when the expression requested by the quiz and the expression of the user do not match, it is regarded as a success. It is not clear that the text that emits the emotion value of the expression of the user analyzed by API is overlapped with the image that is loaded from library. Interworking with mobile. Ui, Ux, interaction, interface design. It was difficult to define ‘what we call ambiguous' needed for the coding building in the team.
To Improve 4&5)Add code so that it shows emotional data without overlapping. Modify the user to experience failure and success in accordance with the quiz text.
Let's design a ranking system so that users can continue to play and be interested : We find ambiguous ambiguity that defines the relationship between ambiguous expressions and emotions as a common language through the ranking and scoring of stored users.