"Build a house with the Three Little Pigs!" "์๊ธฐ๋ผ์ง ์ผํ์ ์ ํจ๊ป ์ง์ ์ง์ด์! ๐ท๐ " "
While child plays, AI screens for ADHD symptoms. ๋์ด์ฒ๋ผ ์ฆ๊ธฐ๋ ์ฌ์ด, AI๊ฐ ADHD ์ฆ์์ ์ดํด๋ด์.
๐ i-There is an AR game for children with ADHD, inspired by The Three Little Pigs. Kids team up with an adorable piglet to gather materials, bake bricks, and build a houseโall within a pastel-toned world featuring intuitive interactions and 3D effect rewards that make play feel like play, not a test. Behind the scenes, AI analyzes gaze, movement, and voice patterns to screen for ADHD symptoms and dynamically adjusts training difficulty to each child's profile.
๐ i-There๋ ์๊ธฐ๋ผ์ง ์ผํ์ ์ด์ผ๊ธฐ๋ฅผ ๋ชจํฐํ๋ก ํ ADHD ์๋์ ์ํ AR ๊ฒ์์ ๋๋ค. ์์ด๋ค์ ๊ท์ฌ์ด ์๊ธฐ ๋ผ์ง ์บ๋ฆญํฐ์ ํจ๊ป ์ฌ๋ฃ๋ฅผ ๋ชจ์ผ๊ณ ๋ฒฝ๋์ ๊ตฌ์ ์ง์ ์์ฑํ๋ ๊ฒ์ ์์์, ํ์คํ ํค ๊ณต๊ฐ๊ณผ ์ง๊ด์ ์ธ ์ธํฐ๋์ , 3D ์ดํํธ ๋ณด์์ ํตํด ๊ฒ์ฌ๋ฐ๋๋ค๋ ๋๋ ์์ด ์ ๋๊ฒ ๋ ์ ์์ต๋๋ค. ๊ทธ ์ฌ์ด AI๊ฐ ์์ด์ ์์ , ์์ง์, ์์ฑ ํจํด ๋ฑ์ ๋ถ์ํด ADHD ์ฆ์์ ์คํฌ๋ฆฌ๋ํ๊ณ ์ฆ์ ์ ๋์ ๋ฐ๋ผ ํ๋ ๋ฐ๋ฌ ํ๋ จ ์ฝํ ์ธ ์ ๋์ด๋๋ฅผ ๊ฐ์ธ๋ณ๋ก ์กฐ์ ํด์ค๋๋ค.
Name: i-There
Genre: Serious Game
Platform: METALENSE2
Duration: 2024.07 ~ 2024.12
| Position | Role | Name | University & Major |
|---|---|---|---|
| ๐ฏ Team Leader | PM & Lead Dev | Jaehyun Byun | Kyung Hee Univ. Computer Science |
| ๐ Planning | Level Designer & Game Planner | Yerim Lee | Ewha Womans Univ. Convergent Contents |
| ๐ Planning | Business Model Planner | Soonmin Kwon | Konkuk Univ. Media Communication |
| ๐ป Dev | Game Client Developer | Seyoung Park | Hongik Univ. Computer Science |
| ๐ป Dev | AI Model Dev & Medical Specialist | Wooky Won | Seoul National Univ. Medicine |
| ๐จ Art | UI/UX Designer | Chowon Jung | Hongik Univ. Painting |
| ๐จ Art | 3D Artist | Yeji Park | Hongik Univ. Ceramics & Glass |
| ๐จ Art | Graphic Designer | Suyeon Lee | Sookmyung Women's Univ. Biological Sciences |
ADHD affects about 7.6% of children worldwide, making early detection and intervention essential. However, current diagnoses rely heavily on subjective assessments, which can feel stressful and burdensome for children. i-There transforms ADHD screening into an engaging AR adventure inspired by The Three Little Pigs, where children build houses while playing. As they play, AI seamlessly collects behavioral data to objectively screen for ADHD symptoms and deliver personalized behavioral development content.
ADHD๋ ์ ์ธ๊ณ ์๋์ ์ฝ 7.6%์๊ฒ ์ํฅ์ ๋ฏธ์น๋ฉฐ, ์กฐ๊ธฐ ์ง๋จ๊ณผ ๊ฐ์ ์ด ์ค์ํฉ๋๋ค. ํ์ง๋ง ํ์ฌ ์ง๋จ์ ์ฃผ๊ด์ ํ๊ฐ์ ์์กดํ๊ณ ์์ด๋ค์ '์ ์ ๊ณผ ๊ฒ์ฌ'๋ผ๋ ๊ณผ์ ์์ ์ฌ๋ฆฌ์ ๋ถ๋ด์ ๋๋ผ๊ธฐ ์ฝ์ต๋๋ค. i-There๋ ์ฆ๊ฐ ํ์ค์ ์๊ธฐ๋ผ์ง ์ผํ์ ์ ์ง์ง๊ธฐ ๋ชจํ ์์์ ์์ด๋ค์ด ์ ๋๊ฒ ๋๋ฉด์, AI๊ฐ ์์ฐ์ค๋ฝ๊ฒ ํ๋ ๋ฐ์ดํฐ๋ฅผ ์์งํด ADHD ์ฆ์์ ๊ฐ๊ด์ ์ผ๋ก ๋ถ์ํ๊ณ ๋ง์ถคํ ํ๋ ๋ฐ๋ฌ ์ฝํ ์ธ ๋ฅผ ์ ๊ณตํ๋ ๊ฒ์ ์๋ฃจ์ ์ ๋๋ค.
Inspired by The Three Little Pigs, kids become the little pig's helperโcollecting clay, baking bricks, and stacking colorful walls to build a house. Child-friendly interface, tactile interactions, gentle voice guidance, and cheerful rewards drive engagement, while the game quietly measures and trains core ADHD cognitive functions.
์๊ธฐ๋ผ์ง ์ผํ์ ์ด์ผ๊ธฐ๋ฅผ ๋ชจํฐํ๋ก, ์์ด๋ค์ ์๊ธฐ ๋ผ์ง์ ์กฐ๋ ฅ์๊ฐ ๋์ด ์งํ์ ๋ชจ์ผ๊ณ , ๋ฒฝ๋์ ๊ตฝ๊ณ , ์๋ก๋ฌ๋กํ ๋ฒฝ์ ์์ ์ง์ ์์ฑํฉ๋๋ค. ์๋์ ์ํ ์ฌ์ด ์ธํฐํ์ด์ค, ์ค์ ์ฌ๋ฌผ์ ๋ค๋ฃจ๋ ๋ฏํ ์ธํฐ๋์ , ๋ค์ ํ ์์ฑ ์๋ด์ ์ ๋๋ ๋ณด์์ด ๋ชฐ์ ์ ์ด๋๋ฉฐ ๊ฒ์์ ๊ทธ ์ฌ์ด ADHD ํต์ฌ ์ธ์ง ๊ธฐ๋ฅ์ ์์ฐ์ค๋ฝ๊ฒ ์ธก์ ํ๊ณ ํ๋ จํฉ๋๋ค.
| Element | Description |
|---|---|
| Art Direction | Storybook-inspired pastel palette, soft lighting, child-friendly rounded silhouettes ๋ํํ ํ์คํ ํ๋ ํธ, ์ํํธ ๋ผ์ดํ , ์๋ ์นํ์ ๋ผ์ด๋ ์ค๋ฃจ์ฃ |
| 3D Modeling | Stylized low-poly pig characters & props in Blender, optimized for XR performance Blender ๊ธฐ๋ฐ ์คํ์ผ๋ผ์ด์ฆ๋ ๋ก์ฐํด๋ฆฌ ์บ๋ฆญํฐยท์ํ, XR ํผํฌ๋จผ์ค ์ต์ ํ |
| Animation | Idle/reaction animations with squash & stretch, particle-based 3D VFX on mission clear ์ค์ฟผ์&์คํธ๋ ์น ์ ์ฉ Idle/๋ฆฌ์ก์ ์ ๋๋ฉ์ด์ , ๋ฏธ์ ํด๋ฆฌ์ด ์ ํํฐํด ๊ธฐ๋ฐ 3D VFX |
| UX Design | Diegetic voice guidance, positive reinforcement on errors, spatial UI with minimap ๋ค์ด์ ํฑ ์์ฑ ์๋ด, ์ค์ ์ ์์ฑ ํผ๋๋ฐฑ, ๋ฏธ๋๋งต ๊ธฐ๋ฐ ๊ณต๊ฐ UI |
Implemented 3-stage mission scenarios (collect clay โ bake bricks โ build house) in Unity. Designed child-friendly UX with Pinch/Grab gesture interactions, minimap-based spatial UI, Naver Clova API voice guidance, and particle-based 3D VFX on mission clear.
3๋จ๊ณ ๋ฏธ์ ์๋๋ฆฌ์ค(์งํ ์์ง โ ๋ฒฝ๋ ๊ตฝ๊ธฐ โ ์ง ์ง๊ธฐ)๋ฅผ Unity์์ ๊ตฌํํ์ต๋๋ค. Pinch/Grab ์ ์ค์ฒ ์ธํฐ๋์ , ๋ฏธ๋๋งต ๊ธฐ๋ฐ ๊ณต๊ฐ UI, Naver Clova API ์์ฑ ์๋ด, ๋ฏธ์ ํด๋ฆฌ์ด ์ ํํฐํด ๊ธฐ๋ฐ 3D VFX๊น์ง ์๋ ์นํ์ UX๋ฅผ ์ค๊ณํ์ต๋๋ค.
Converted DSM-5 ADHD diagnostic criteria (18 items) into digital behavioral metrics. CNN model learns temporal patterns from behavioral sequences. PyTorch โ ONNX conversion enables on-device real-time inference via Unity Barracuda. Inference results (0~54 score) dynamically adjust Mission 3 difficulty.
DSM-5 ๊ธฐ๋ฐ ADHD ์ง๋จ ๊ธฐ์ค 18๊ฐ ํญ๋ชฉ์ ๋์งํธ ํ๋ ์งํ๋ก ๋ณํํ์ต๋๋ค. CNN ๋ชจ๋ธ์ด ํ๋ ์ํ์ค์์ ์๊ณ์ด ํจํด์ ํ์ตํ๊ณ , tensorflow โ ONNX ๋ณํ ํ Unity Barracuda๋ก ์จ๋๋ฐ์ด์ค ์ค์๊ฐ ์ถ๋ก ์ ์ํํฉ๋๋ค. ์ถ๋ก ๊ฒฐ๊ณผ(0~54์ )์ ๋ฐ๋ผ Mission 3์ ๋์ด๋๊ฐ ๋์ ์ผ๋ก ์กฐ์ ๋ฉ๋๋ค.
๐ Reference: A Novel Approach to Diagnose ADHD Using Virtual Reality
Collects multi-dimensional behavioral data (gaze, hand movement, head rotation) from hand/eye tracking and IMU sensors. Memory-efficient processing via Unity Static arrays, followed by normalization and filtering. REST API transmits to server, MariaDB stores per-user screening results and session logs for long-term tracking and model improvement.
ํธ๋/์์ด ํธ๋ํน๊ณผ IMU ์ผ์์์ ์์ , ์ ์์ง์, ๋จธ๋ฆฌ ํ์ ๋ฑ ๋ค์ฐจ์ ํ๋ ๋ฐ์ดํฐ๋ฅผ ์์งํฉ๋๋ค. Unity Static ๋ฐฐ์ด๋ก ๋ฉ๋ชจ๋ฆฌ ํจ์จ์ ์ฒ๋ฆฌ ํ ์ ๊ทํ ๋ฐ ํํฐ๋ง์ ๊ฑฐ์ณ REST API๋ก ์๋ฒ์ ์ ์ก, MariaDB์ ์ฌ์ฉ์๋ณ ์คํฌ๋ฆฌ๋ ๊ฒฐ๊ณผ์ ์ธ์ ๋ก๊ทธ๋ฅผ ์ ์ฅํ์ฌ ์ฅ๊ธฐ ์ถ์ ๋ฐ ๋ชจ๋ธ ๊ฐ์ ์ ํ์ฉํฉ๋๋ค.
โโโโโโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโโโโโ
โ XR Device โ โ Unity Client Engine โ โ AI Inference โ โ Backend โ
โ METALENSE2 โ โ โ โ โ โ โ
โโโโโโโโโโโโโโโโโโโโค โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค โโโโโโโโโโโโโโโโโโโโค โโโโโโโโโโโโโโโโโโโโค
โ โข Hand Tracking โ โ โโโโโโโโโโโโ โโโโโโโโโโโโ โโโโโโโโ โ โ Barracuda CNN โ โ REST API Server โ
โ โข Eye Tracking โโโโโโ โMission 1 โโโMission 2 โโโ M3 โ โโโโโโ โโโโโโ โ
โ โข IMU Sensor โ โ โClay โ โBricks โ โBuild โ โ โ DSM-5 18-Item โ โ MariaDB โ
โ โข Voice Input โ โ โCollectionโ โBaking โ โHouse โ โ โ Scoring (0~54) โ โ User Profiles โ
โ โข Gameplay Data โ โ โโโโโโโโโโโโ โโโโโโโโโโโโ โโโโฌโโโโ โ โ โ โ โ Session Logs โ
โ (Time/Errors) โ โ โ โ โ โ โ โ Long-term Track โ
โโโโโโโโโโโโโโโโโโโโ โ Adaptive Difficulty โ โโโโโโโโโโโโโโโโโโโโ โโโโโโโโโโโโโโโโโโโโ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ ๏ธ This game was developed using Metacore SDK and runs exclusively on METALENSE2 XR glasses.
Presented at the XR Device Content Makeathon finals. During live demonstration where judges played hands-on, the full systemโgameplay flow, real-time AI screening, and server integrationโran stable, proving the playable build's completeness. Awarded 2nd place with NIPA Director's Award and โฉ5M prize.
๊ณผํ๊ธฐ์ ์ ๋ณดํต์ ๋ถ ์ฃผ์ต XR ๋๋ฐ์ด์ค ์ฝํ ์ธ ๋ฉ์ด์ปคํค ๋ณธ์ ๋ฐํ. ์ฌ์ฌ์์๋ค์ด ์ง์ ํ๋ ์ดํ ํ์ฅ ์์ฐ์์ ๊ฒ์ ํ๋ ์ด, ์ค์๊ฐ AI ์คํฌ๋ฆฌ๋, ์๋ฒ ์ฐ๋๊น์ง ์ ์์คํ ์ด ์์ ์ ์ผ๋ก ๊ตฌ๋๋๋ฉฐ ํ๋ ์ด์ด๋ธ ๋น๋์ ์์ฑ๋๋ฅผ ์ ์ฆํ์ต๋๋ค. ์ ๋ณดํต์ ์ฐ์ ์งํฅ์์ฅ์๊ณผ ์๊ธ 500๋ง์ ์์.
Built on i-There's tech foundation, refined design via expert interviews (child development professor, psychiatry resident, digital healthcare investor). Redesigned around The Little Princeโ3-stage screening (conversing with Prince, caring for Rose, collecting fuel) plus personalized behavioral missions on thematic planets. Narrative drives self-efficacy through bonding and self-discovery; ADHD-optimized visuals reduce cognitive load. Competed as design proposal and prototype against Georgia Tech, UCLA, UCL, Shanghai Jiao Tongโselected as Finalist (top 4).
i-There ๊ธฐ์ ๊ธฐ๋ฐ ์์ ์๋ํ ๊ต์, ์ ์ ๊ณผ ์ ๊ณต์, ๋์งํธ ํฌ์ค์ผ์ด ํฌ์์ฌ ๋ํ ์ธํฐ๋ทฐ๋ฅผ ํตํด ๊ฒ์ ๋์์ธ ๊ฐ์ . '์ด๋ฆฐ ์์'๋ฅผ ๋ชจํฐํ๋ก ์ฌ์ค๊ณโ์ด๋ฆฐ ์์์์ ๋ํ, ์ฅ๋ฏธ ๋๋ณด๊ธฐ, ๋นํ ์ฐ๋ฃ ์์ง์ 3๋จ๊ณ ์คํฌ๋ฆฌ๋ ์ํ์ค์ ์์ ํ์ฑ, ์์ธ์ ํ์ฑ ๋ฑ ๊ฐ์ธํ๋ ํ๋ ๋ฐ๋ฌ ๋ฏธ์ ๊ตฌ์ฑ. ์ ๋์ ์๊ธฐ ๋ฐ๊ฒฌ์ ์์ฌ๋ก ๋ถ์ ์ ์์ ์ธ์ ๊ทน๋ณต๊ณผ ์๊ธฐ ํจ๋ฅ๊ฐ ํ๋ณต์ ์ ๋ํ๊ณ , ์ฅํ์ฅ ์์๊ณผ ์ง๊ด์ ๊ทธ๋ํฝ์ผ๋ก ADHD ์๋ ํนํ ์๊ฐ ๋์์ธ ๊ฐ์ . ๊ธฐํ์๊ณผ ๋์์ธ ํ๋กํ ํ์ ์ผ๋ก ์ ๋ฆฌํ์ฌ ๋ณธ์ ์ง์ถ ํ Georgia Tech, UCLA, UCL, ์ํ์ด ๊ตํต๋ํ๊ต ๋ฑ๊ณผ ๊ฒฝ์, ์ต์ข Finalist (Top 4) ์ ์ ์ ๋์์ต๋๋ค.











