This week is the final development week before the soft opening. We really made huge progress this week. First, we finished the development of the reinforcement learning prototype and the VR painting prototype. We also conducted three playtests this week on the music prototype, VR painting prototype, and reinforcement learning prototype separately. Also, we decided to submit our project to SIGGRAPH as a poster. We finished our first draft for the conference.
Playtest of Music Prototype
This Monday, we had our third playtest for the music prototype. Based on the last two playtest results, we found that the Novelty of interaction can spark curiosity. However, the novelty may be so overwhelming that it hinders the learning process. Therefore, we decided to conduct our third playtest to invite ETC students who tried this prototype on our first playtest. This time, our goal is to test if the user can understand the music data using haptic feeling. The playtest rule is the same as last time. We provided 3 songs in total for this playtest. For all three songs, the haptic feeling is matched with the visualization. However, only the first song has the matched music. For the second and third songs, what the user listens to is different from what they see and what they feel. The second song is slightly disconnected. The third song is totally disconnected. During the playtest, we didn’t tell the playtester the song was not matched.
The playtest result is really interesting. There are 9 playtesters in total. During the playtest time, 4 playtesters mentioned that the beat or note is not matched with the music, which really surprised us. The below diagram shows our playtest result. The question is Do you feel that the haptic feeling represented this song well. And as you can see, the score of the first song is much higher than the second and third songs.
So finally, we got our conclusion of this prototype from the three playtests results.
- The novelty of interaction can spark curiosity.
- However, the novelty may be overwhelming that it hinders the learning process.
- Experiencing multiple times allow people to learn more
- Engaging experiences can allow people to be inclined to explore the area of the “unknown” and build a cyclical relationship between the pursuit of learning and engagement.
- Haptic feedback can represent a specific song. It can be a unique way of experiencing music.
VR Painting Prototype & Playtest
This week, we finished the VR Painting Prototype. We bring our Sprint 1 to the VR platform. So the user can explore a gallery and see how different paintings are split into different layers using the k-means algorithm. The user can also walk into the painting and also adjust the size of the painting, rotation speed, and distance between each layer. So let’s see the demo video first.
We had a playtest for this prototype on Wednesday. The main goal of this playtest is to compare the difference between PC and VR. We want to see on which platform the user understands the data better and on which platform the user is more engaging. So during the playtest, we provided VR and PC prototypes at the same time, the user needed to try both prototypes and fill in a questionnaire.
In total there were 17 ETC student participants in the playtest. More than 70% of playtesters think that the VR prototype is more engaging than the PC prototype. Also, from our observation, we found that players spend more time on the VR prototype compared to the PC prototype. For the understanding of data, most playtesters think there is no huge difference between the two platforms. So here is our conclusion.
- Immersive Experience in VR Platform improves users’ interests
- Users are willing to spend more time discovering the painting in VR space
- There was no significant difference in the level of understanding of data between the Desktop / VR
Reinforcement Learning Prototype & Playtest
Last week, we already finished most part of the reinforcement learning prototype. So this Monday, we presented our prototype to our instructor Ruth and she really provided us with a lot of useful feedback and suggestions. So we refined our prototype this week and conducted the playtest on Friday. We add the animation of the dinosaur. Also, we added sound effects and background music. We redesigned the story and changed some UI elements. So here is our demo video.
We conducted a playtest on Friday and 8 ETC students participated in the playtest. So first we want to test if the user can understand the basic concept of Reinforcement Learning through this experience. Also, we want to test if storytelling can help the user understand the algorithm better.
Here is our result. First, all the playtesters can understand that dinosaurs stand for AI, meat stands for the goal and lava stands for punishment. And about 30% of players feel a little bit confused about what the rock stands for. And all the playtesters think that having dinosaur/meat/lava/rock as visual metaphors helped better understand the reinforcement learning. And 75% of playtesters think that storytelling can help them understand the algorithm better. One interesting finding is that players really enjoy our sandbox mode. Many of them spend more than 10 minutes in sandbox mode. According to the feedback, they said the sandbox is the part they like most because they can create whatever they want. So here are some screenshots of the map created by the playtesters.
We will invite more people to playtest next week. And here is our brief conclusion based on the current result of this prototype.
- Providing storytelling with relative metaphors can help users to be inclined to learn
- Gamifying experience by rewarding and punishing behaviors improves user engagement and has a higher retention rate
- Providing levels focusing on different functions of RL can foster a better understanding of the possibilities and constraints of a machine learning model
SIGGRAPH Poster
We decided to submit our project as a poster to SIGGRAPH. So we finished a draft of the documentation this week. We will share it with our instructors and client to ask for feedback next week.