Skip to content

abbosaliboev/ICCAS_2026

Folders and files

NameName
Last commit message
Last commit date

Latest commit

ย 

History

251 Commits
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 
ย 

Repository files navigation

MobiCare

Edge AI-Based Real-Time Fall Detection and Alert System for Elderly People Living Alone

๋…๊ฑฐ ๋…ธ์ธ์„ ์œ„ํ•œ ์—ฃ์ง€ AI ๊ธฐ๋ฐ˜ ์‹ค์‹œ๊ฐ„ ๋‚™์ƒ ๊ฐ์ง€ ๋ฐ ์•Œ๋ฆผ ์‹œ์Šคํ…œ

๐Ÿ“„ Poster: Docs/Poster/Team-4_Poster_Final.jpg

MobiCare Poster (EKC 2026)

๐Ÿ“„ Presentation: Docs/PPT/260713_PPT_FinalFinal.pdf

Language / ์–ธ์–ด ์„ ํƒ


โšก Quick Start / ๋น ๋ฅธ ์‹œ์ž‘ / Tezkor Boshlash

Full guide โ†’ fall_iccas/HOW_TO_RUN.md

1. Backend (PC)

cd backend && python main.py
# Web app: http://localhost:8000/app

2. Edge server (Jetson โ€” RTSP camera + TensorRT)

cd fall_iccas
python edge_server.py \
  --source "rtsp://admin:PASSWORD@CAMERA_IP:554/stream1" \
  --backend "http://PC_IP:8000" \
  --tensorrt
# MJPEG stream: http://JETSON_IP:8081/video

3. Training (PC with GPU)

cd fall_iccas
python prepare_cv_dataset.py   # extract keypoints from UP-Fall images
python train_two_stage.py      # train ST-GCN + physics filter

๐Ÿ“„ Documents

Document Link
All Features fall_iccas/FEATURES.md
How to Run fall_iccas/HOW_TO_RUN.md
Architecture fall_iccas/ARCHITECTURE.md
Training Results fall_iccas/RESULTS.md
FPS Benchmark fall_iccas/FPS_BENCHMARK.md
Limitations fall_iccas/LIMITATIONS.md
Abstract (Markdown) Docs/Poster/abstract.md
Poster (PDF) Docs/Poster/4์กฐ ์ดˆ๋ก.pdf

๐ŸŽฅ Demo โ€” Fall Detection (ST-GCN + Physics Filter)

Real-time webcam test: ST-GCN model combined with the physics rescue filter.

Fall Detection Demo


๐Ÿ—“๏ธ Conference Schedule / ํ•™์ˆ ๋Œ€ํšŒ ์ผ์ •

Toulouse, France โ€” July 2026

Event Description Date
EKC 2026 18th Europe-Korea Conference on Science and Technology โ€” poster presentation July 20โ€“23, 2026
ICCAS 2026 International Collegiate Challenge for AI-Assisted Society โ€” paper presentation July 18โ€“24, 2026

EKC (Europe-Korea Conference) is an annual forum organized by KOFST that connects Korean scientists and engineers across Europe and Korea. The 2026 edition is held in Toulouse, France.

Team departs from Chungbuk National University (CBNU), South Korea ยท July 6โ€“15, 2026.


ํ”„๋ž‘์Šค ํˆด๋ฃจ์ฆˆ โ€” 2026๋…„ 7์›”

ํ–‰์‚ฌ ์„ค๋ช… ์ผ์ •
EKC 2026 ์ œ18ํšŒ ํ•œ-์œ ๋Ÿฝ ๊ณผํ•™๊ธฐ์ˆ ํ•™์ˆ ๋Œ€ํšŒ โ€” ํฌ์Šคํ„ฐ ๋ฐœํ‘œ 2026๋…„ 7์›” 20โ€“23์ผ
ICCAS 2026 International Collegiate Challenge for AI-Assisted Society โ€” ๋…ผ๋ฌธ ๋ฐœํ‘œ 2026๋…„ 7์›” 18โ€“24์ผ

EKC(ํ•œ-์œ ๋Ÿฝ ๊ณผํ•™๊ธฐ์ˆ ํ•™์ˆ ๋Œ€ํšŒ)๋Š” KOFST๊ฐ€ ์ฃผ๊ด€ํ•˜๋Š” ์—ฐ๋ก€ ํฌ๋Ÿผ์œผ๋กœ, ์œ ๋Ÿฝ๊ณผ ํ•œ๊ตญ์˜ ๊ณผํ•™๊ธฐ์ˆ ์ธ์„ ์—ฐ๊ฒฐํ•ฉ๋‹ˆ๋‹ค. 2026๋…„ ๋Œ€ํšŒ๋Š” ํ”„๋ž‘์Šค ํˆด๋ฃจ์ฆˆ์—์„œ ๊ฐœ์ตœ๋ฉ๋‹ˆ๋‹ค.

ํŒ€ ์ถœ๋ฐœ: ์ถฉ๋ถ๋Œ€ํ•™๊ต (CBNU) ยท 2026๋…„ 7์›” 6โ€“15์ผ.


๐Ÿ‡บ๐Ÿ‡ธ English

1. What Service?

Core Message

"Stop it before it happens. Detect it the moment it does. Bring them back after."


Who Is It For?

Korea and France have both entered super-aged societies โ€” people aged 65 and older now make up over 20% of the total population. In France, about 30% of people aged 65+ experience a fall every year. In Korea, falls account for the highest proportion of all injury causes.

MobiCare primarily targets elderly people living alone, where no caregiver is present to respond immediately after a fall.

Target Scale Core Risk
Elderly living alone (65+) ~2 million in Korea Left unattended after a fall โ€” no one to respond
Nursing hospital patients High-risk environment Repeated falls, staff cannot monitor all patients
Senior welfare facility residents Institutional setting Falls during nighttime or unsupervised hours

Where Does It Work?

All indoor spaces:
โœ… Living room / Hallway / Bedroom / Kitchen
โœ… Stairs (the #1 location for fall accidents)
โœ… Yard / Garden / Balcony
โŒ Bathroom / Restroom (privacy)

Top 3 fall risk zones: living room, bathroom surroundings, bedroom โ€” each with different characteristics and risk patterns.


How It Works โ€” 3 Stages

โ‘  PREVENT
   Daily "Frozen!" game โ€” 5 minutes
   โ†’ AI quietly measures balance ability
   โ†’ "Balance declined 15% this week" โ†’ early warning sent to family

โ‘ก DETECT
   24/7 camera + audio analysis
   โ†’ Fall detected instantly
   โ†’ Family app notification + SMS within 3 seconds

โ‘ข REHABILITATE
   Personalized exercise program after a fall
   โ†’ YOLOv11 analyzes posture and movement in real time
   โ†’ Weekly rehabilitation report โ†’ sent automatically to hospital

2. What Product?

Limitations of Existing Solutions

Existing Product Limitation
Smartwatch fall detection Detects after the fall. No prevention. High resistance to wear among elderly
Home CCTV Recording only. No AI analysis. No alerts
Hospital rehabilitation programs 1โ€“2 visits per week. Requires travel. No daily monitoring
Fall detection mats Detects only one spot. Not portable. No rehab function

All existing wearable sensor-based systems require users to wear a device at all times โ€” uncomfortable, and they cannot track full-body movement since they only measure one body part.

What MobiCare Improves

Before:   fall happens โ†’ detected
MobiCare: before fall โ†’ predict โ†’ prevent
          fall happens โ†’ instant detection โ†’ alert
          after fall   โ†’ rehabilitation โ†’ recovery

Only two things needed:

  • Regular camera (webcam or IP camera)
  • Smartphone (family app)

No extra sensors, wearables, or special equipment required.


Core Technology

๐Ÿ“ท Single Camera Input
    โ†“
Edge AI Device (Lightweight Model)
    โ†“
YOLOv11 Pose Estimation
โ†’ Real-time tracking of 17 body joints
โ†’ Fall pattern detection
โ†’ Balance Score quantification
    โ†“
Audio Analysis Model
โ†’ Crying / "Help!" / Impact sound detection
    โ†“
Fall Decision Logic
โ†’ Minimize false alarms
   (distinguish sitting down vs. actual fall)
    โ†“
Alert Service (Push / SMS / KakaoTalk)
    โ†“
Family app + Dashboard

Edge AI: the model runs directly on an edge device โ€” low latency, no dependency on cloud connection, privacy-safe.


Game: "Frozen!" โ€” Prevention as a Game

Balance training from the clinic โ€” now at home, every day, and fun.

  • Simple game: when the music stops, freeze in place
  • YOLOv11 measures stability of the frozen pose in real time
  • Anyone in the world understands in 5 seconds โ€” no language barrier
  • Ranking system for ongoing motivation
Ranking structure:
โ”œโ”€โ”€ Personal ranking (last week me vs. this week me)
โ”œโ”€โ”€ Local ranking (how do I rank in my neighborhood?)
โ””โ”€โ”€ Age group ranking (where do I rank among people in their 70s?)

Game Interaction โ€” Two Options Under Review

One will be chosen, or decided based on implementation difficulty.

Option A โ€” Free Movement Mode

Music plays โ†’ user moves freely
Music stops โ†’ freeze in place
AI measures only stability of the frozen pose
โ†’ Any pose is fine โ€” score is based on how still you are
  • Pros: easy to implement, anyone can do it instantly, no instructions needed
  • Cons: hard to standardize as a rehabilitation exercise

Option B โ€” Guided Pose Mode

Music plays โ†’ target pose (template) shown on screen
User follows the pose
Music stops โ†’ freeze in that pose
AI measures pose accuracy + stability simultaneously
  • Pros: clear rehabilitation effect, posture correction possible
  • Cons: complex to implement (pose matching logic), difficult for users who struggle to follow instructions

Current decision:

If template is too complex โ†’ Option A (free movement)
If feasible โ†’ Option B (guided) or both modes combined

Personalized Game Mode

Identify who the user is first. Then set game conditions automatically.

On first launch, the user sets up a profile.
Game difficulty, pose type, and scoring criteria are automatically personalized.

User profile setup
โ”œโ”€โ”€ Age / Gender
โ”œโ”€โ”€ Current condition:
โ”‚   โ”œโ”€โ”€ General elderly (balance maintenance)
โ”‚   โ”œโ”€โ”€ Stroke rehabilitation patient
โ”‚   โ”œโ”€โ”€ Dementia patient
โ”‚   โ”œโ”€โ”€ Post-orthopedic surgery recovery
โ”‚   โ””โ”€โ”€ Person with physical disability (wheelchair / walking aid)
โ””โ”€โ”€ Rehabilitation goal (balance / fall prevention / strength recovery)

Game parameters by user type:

User Type Music Tempo Freeze Duration Tracked Area Sway Tolerance
General elderly Normal 5 sec Full body Normal
Stroke patient Slow 8 sec Upper body focus Wide
Dementia patient Slow + simple 10 sec Full body (simple poses) Wide
Post-surgery recovery Very slow 10 sec Minimize lower body strain Wide
Physical disability Normal 5 sec Seated pose standard Upper body only

Automatic difficulty adjustment over time:

Week 1: easy poses, slow music, long freeze time
       โ†“ if Balance Score improves
Week 2: automatically upgraded to slightly harder poses
       โ†“ if no improvement
       โ†’ alert sent to family

3. How Do We Prove It Works?

Measurable Metrics (Data-Driven)

โ‘  Balance Score

Measurement method:
YOLOv11 โ†’ tracks Center of Mass (CoM)
โ†’ measures sway range per second (cm)
โ†’ quantified as 0โ€“100 score

Clinical reference: validated against Berg Balance Scale (BBS)

โ‘ก Fall Detection Accuracy

Precision / Recall / F1-Score
Target: Precision 90%+, Recall 95%+
(minimize false alarms, never miss a real fall)

Baseline: standard video-based action recognition model

โ‘ข Rehabilitation Progress

Weekly Balance Score change rate
Exercise consistency (consecutive days participated)
Reduction in hospital visit frequency

โ‘ฃ Alert Response Time

Fall event โ†’ family notification received:
Target: within 3 seconds

Data Collection Types

Skeleton data:
  - 17 joints (x, y, confidence score)
  - 30 FPS real-time tracking

Audio data:
  - MFCC features
  - 16kHz sampling rate
  - Classification: crying / impact sound / "Help!"

Balance Score:
  - CoM displacement (cm/s)
  - Delta from session start to end

Event log:
  - timestamp, room ID, alert type, response time

Dataset strategy: Initial model validation using Le2i Fall Detection Dataset + UR Fall Detection Dataset.
Followed by small-scale in-home data collection in real environments.


Data Visualization โ€” Dashboard

Family app:
โ”œโ”€โ”€ Real-time alert (when fall occurs)
โ”œโ”€โ”€ Today's activity summary
โ””โ”€โ”€ Weekly health trend graph

Hospital / Doctor view:
โ”œโ”€โ”€ Weekly Balance Score change
โ”œโ”€โ”€ Exercise participation rate
โ”œโ”€โ”€ Fall history log
โ””โ”€โ”€ AI-generated rehabilitation report (Gemini AI)

Danger zone heatmap:
โ””โ”€โ”€ Top 3 risk zones in the home โ€” living room, bedroom, hallway
    (each zone has different fall characteristics โ†’ different response)

4. Expected Impact

Individual Level

Before MobiCare:
fall โ†’ no one knows โ†’ found hours later โ†’ fracture โ†’ hospitalization โ†’ social isolation

After MobiCare:
balance decline detected โ†’ rehab exercise โ†’ fall prevented โ†’ mobility maintained
                    OR
fall detected instantly โ†’ family notified in 3 seconds โ†’ rapid response
  • Fall rate reduction (research basis: 30โ€“40% reduction with consistent rehab exercise)
  • Time to discovery after a fall: average 1 hour โ†’ 3 seconds
  • Rehabilitation participation: hospital-based 1โ€“2x/week โ†’ daily possible

Social Level

Aging society + rural community decline
        โ†“
Growing number of elderly living alone
        โ†“
MobiCare
        โ†“
โ‘  Reduced medical costs (fall prevention = fewer hospitalizations)
โ‘ก Reduced psychological burden on families
โ‘ข Extended period of independent living for elderly
โ‘ฃ Applicable to nursing hospitals and senior welfare facilities

Based on Korea's National Health Insurance Service data,
fall-related hospitalization costs are estimated at over 1 trillion KRW per year.
MobiCare can structurally reduce a portion of this cost.


The Meaning in Terms of AI Mobility

"Mobility is not just about walking.
It's about participating in society, staying connected to family, and living with dignity.
MobiCare protects that ability with AI."

MobiCare is not a fall detection app.
It is a platform that uses Edge AI to protect the lives of elderly people living alone.


Summary

Question Answer
What? Edge AI fall detection + balance game + alert system
Who? Elderly living alone, nursing hospital & welfare facility residents
When? 24/7 โ€” whenever the elderly person is without a caregiver
Where? Living room, bedroom, stairs, hallway
Why? Falls among the elderly living alone go unnoticed for hours
How? Single camera + lightweight Edge AI + instant alert (SMS/Push)


๐Ÿ‡ฐ๐Ÿ‡ท ํ•œ๊ตญ์–ด

1. ์–ด๋–ค ์„œ๋น„์Šค๋ฅผ?

ํ•ต์‹ฌ ํ•œ ์ค„

"๋„˜์–ด์ง€๊ธฐ ์ „์— ๋ง‰๊ณ , ๋„˜์–ด์ง€๋ฉด ์ฆ‰์‹œ ์•Œ๋ฆฌ๊ณ , ๋„˜์–ด์ง„ ํ›„์—” ๋‹ค์‹œ ์ผ์œผํ‚จ๋‹ค."


๋ˆ„๊ตฌ๋ฅผ ์œ„ํ•œ ์„œ๋น„์Šค์ธ๊ฐ€?

ํ•œ๊ตญ๊ณผ ํ”„๋ž‘์Šค๋Š” ๋ชจ๋‘ ์ดˆ๊ณ ๋ น์‚ฌํšŒ์— ์ง„์ž…ํ–ˆ์Šต๋‹ˆ๋‹ค โ€” 65์„ธ ์ด์ƒ ์ธ๊ตฌ๊ฐ€ ์ „์ฒด์˜ 20% ์ด์ƒ์ž…๋‹ˆ๋‹ค.
ํ”„๋ž‘์Šค์—์„œ๋Š” 65์„ธ ์ด์ƒ ๋…ธ์ธ์˜ ์•ฝ **30%**๊ฐ€ ๋งค๋…„ ๋‚™์ƒ์„ ๊ฒฝํ—˜ํ•˜๋ฉฐ,
ํ•œ๊ตญ์—์„œ๋„ ๋‚™์ƒ์€ ์ „์ฒด ์†์ƒ ์›์ธ ์ค‘ ๊ฐ€์žฅ ๋†’์€ ๋น„์ค‘์„ ์ฐจ์ง€ํ•ฉ๋‹ˆ๋‹ค.

MobiCare๋Š” ๋…๊ฑฐ ๋…ธ์ธ์„ ์ฃผ์š” ๋Œ€์ƒ์œผ๋กœ ํ•ฉ๋‹ˆ๋‹ค โ€” ๋‚™์ƒ ํ›„ ์ฆ‰๊ฐ์ ์ธ ํ™•์ธ์ด ์–ด๋ ค์šด ํ™˜๊ฒฝ์— ์žˆ๋Š” ๋ถ„๋“ค์ž…๋‹ˆ๋‹ค.

๋Œ€์ƒ ๊ทœ๋ชจ ํ•ต์‹ฌ ์œ„ํ—˜
๋…๊ฑฐ ๋…ธ์ธ (65์„ธ+) ํ•œ๊ตญ ์•ฝ 200๋งŒ ๋ช… ๋‚™์ƒ ํ›„ ์žฅ์‹œ๊ฐ„ ๋ฐฉ์น˜ โ€” ์ฆ‰๊ฐ ๋Œ€์‘ ๋ถˆ๊ฐ€
์š”์–‘๋ณ‘์› ํ™˜์ž ๊ณ ์œ„ํ—˜ ํ™˜๊ฒฝ ๋ฐ˜๋ณต ๋‚™์ƒ, ์ „์ฒด ํ™˜์ž ์ƒ์‹œ ๊ฐ์‹œ ๋ถˆ๊ฐ€
๋…ธ์ธ๋ณต์ง€์‹œ์„ค ๊ฑฐ์ฃผ์ž ์‹œ์„ค ํ™˜๊ฒฝ ์•ผ๊ฐ„ ๋˜๋Š” ๋ฌด๊ฐ๋… ์‹œ๊ฐ„๋Œ€ ๋‚™์ƒ

์–ด๋–ค ๊ณต๊ฐ„์—์„œ?

์ง‘ ์•ˆ ๋ชจ๋“  ๊ณต๊ฐ„:
โœ… ๊ฑฐ์‹ค / ๋ณต๋„ / ์นจ์‹ค / ์ฃผ๋ฐฉ
โœ… ๊ณ„๋‹จ (๋‚™์ƒ ์‚ฌ๊ณ  1์œ„ ์žฅ์†Œ)
โœ… ๋งˆ๋‹น / ์ •์› / ๋ฐœ์ฝ”๋‹ˆ
โŒ ์š•์‹ค / ํ™”์žฅ์‹ค (ํ”„๋ผ์ด๋ฒ„์‹œ)

๋‚™์ƒ ์œ„ํ—˜ ๊ตฌ์—ญ Top 3: ๊ฑฐ์‹ค, ์š•์‹ค ์ฃผ๋ณ€, ์นจ์‹ค โ€” ๊ฐ ๊ตฌ์—ญ๋งˆ๋‹ค ๋‚™์ƒ ํŠน์„ฑ๊ณผ ๋Œ€์‘ ๋ฐฉ์‹์ด ๋‹ค๋ฆ…๋‹ˆ๋‹ค.


์„œ๋น„์Šค ๋ฐฉ๋ฒ• โ€” 3๋‹จ๊ณ„

โ‘  PREVENT (์˜ˆ๋ฐฉ)
   ๋งค์ผ "Frozen!" ๊ฒŒ์ž„ 5๋ถ„
   โ†’ AI๊ฐ€ ๊ท ํ˜• ๋Šฅ๋ ฅ์„ ์กฐ์šฉํžˆ ์ธก์ •
   โ†’ "์ด๋ฒˆ ์ฃผ ๊ท ํ˜• ๋Šฅ๋ ฅ 15% ์ €ํ•˜" โ†’ ๊ฐ€์กฑ์—๊ฒŒ ์‚ฌ์ „ ๊ฒฝ๊ณ 

โ‘ก DETECT (๊ฐ์ง€)
   24์‹œ๊ฐ„ ์นด๋ฉ”๋ผ + ์Œ์„ฑ ๋ถ„์„
   โ†’ ๋‚™์ƒ ๋ฐœ์ƒ ์ฆ‰์‹œ ๊ฐ์ง€
   โ†’ 3์ดˆ ์•ˆ์— ๊ฐ€์กฑ ์•ฑ ์•Œ๋ฆผ + ๋ฌธ์ž

โ‘ข REHABILITATE (์žฌํ™œ)
   ๋‚™์ƒ ํ›„ ๋งž์ถค ์šด๋™ ํ”„๋กœ๊ทธ๋žจ
   โ†’ YOLOv11์ด ์ž์„ธ์™€ ๋™์ž‘์„ ์‹ค์‹œ๊ฐ„ ๋ถ„์„
   โ†’ ์ฃผ๊ฐ„ ์žฌํ™œ ๋ฆฌํฌํŠธ โ†’ ๋ณ‘์› ์ž๋™ ์ „์†ก

2. ์–ด๋–ค ์ œํ’ˆ์œผ๋กœ?

๊ธฐ์กด ์œ ์‚ฌ ์ œํ’ˆ์˜ ํ•œ๊ณ„

๊ธฐ์กด ์ œํ’ˆ ํ•œ๊ณ„์ 
์Šค๋งˆํŠธ์›Œ์น˜ ๋‚™์ƒ ๊ฐ์ง€ ๋„˜์–ด์ง„ ํ›„ ๊ฐ์ง€. ์˜ˆ๋ฐฉ ๋ถˆ๊ฐ€. ๊ณ ๋ น์ž ์ฐฉ์šฉ ๊ฑฐ๋ถ€๊ฐ ๋†’์Œ
๊ฐ€์ •์šฉ CCTV ๋‹จ์ˆœ ๋…นํ™”๋งŒ. AI ๋ถ„์„ ์—†์Œ. ์•Œ๋ฆผ ์—†์Œ
๋ณ‘์› ์žฌํ™œ ํ”„๋กœ๊ทธ๋žจ ์ฃผ 1~2ํšŒ, ๋ณ‘์› ๋ฐฉ๋ฌธ ํ•„์ˆ˜. ์ผ์ƒ ๋ชจ๋‹ˆํ„ฐ๋ง ๋ถˆ๊ฐ€
๊ธฐ์กด ๋‚™์ƒ ๊ฐ์ง€ ๋งคํŠธ ํŠน์ • ์œ„์น˜๋งŒ ๊ฐ์ง€. ์ด๋™ ๋ถˆ๊ฐ€. ์žฌํ™œ ๊ธฐ๋Šฅ ์—†์Œ

๊ธฐ์กด ์›จ์–ด๋Ÿฌ๋ธ” ์„ผ์„œ ๋ฐฉ์‹์€ ์‚ฌ์šฉ์ž๊ฐ€ ์žฅ์น˜๋ฅผ ํ•ญ์ƒ ์ฐฉ์šฉํ•ด์•ผ ํ•˜๋ฉฐ, ๋‹จ์ผ ๋ถ€์œ„ ์ธก์ •์œผ๋กœ ์ „์‹  ๊ฑฐ๋™ ํŒŒ์•…์ด ๋ถˆ๊ฐ€ํ•ฉ๋‹ˆ๋‹ค.

MobiCare๊ฐ€ ๊ฐœ์„ ํ•˜๋Š” ๊ฒƒ

๊ธฐ์กด:  ๋„˜์–ด์ง„ ํ›„ โ†’ ๊ฐ์ง€
MobiCare: ๋„˜์–ด์ง€๊ธฐ ์ „ โ†’ ์˜ˆ์ธก โ†’ ์˜ˆ๋ฐฉ
           ๋„˜์–ด์ง„ ํ›„  โ†’ ์ฆ‰์‹œ ๊ฐ์ง€ โ†’ ์•Œ๋ฆผ
           ๋„˜์–ด์ง„ ๋‹ค์Œ โ†’ ์žฌํ™œ โ†’ ํšŒ๋ณต

ํ•„์š”ํ•œ ๊ฒƒ์€ ๋‹จ ๋‘ ๊ฐ€์ง€:

  • ์ผ๋ฐ˜ ์นด๋ฉ”๋ผ (์›น์บ  ๋˜๋Š” IP ์นด๋ฉ”๋ผ)
  • ์Šค๋งˆํŠธํฐ (๊ฐ€์กฑ์šฉ ์•ฑ)

์ถ”๊ฐ€ ์„ผ์„œ, ์›จ์–ด๋Ÿฌ๋ธ”, ํŠน์ˆ˜ ์žฅ๋น„ โ€” ์ „ํ˜€ ํ•„์š” ์—†์Šต๋‹ˆ๋‹ค.


ํ•ต์‹ฌ ๊ธฐ์ˆ  ๊ตฌ์„ฑ

๐Ÿ“ท ๋‹จ์ผ ์นด๋ฉ”๋ผ ์ž…๋ ฅ
    โ†“
์—ฃ์ง€ AI ๋””๋ฐ”์ด์Šค (๊ฒฝ๋Ÿ‰ํ™” ๋ชจ๋ธ)
    โ†“
YOLOv11 Pose Estimation
โ†’ 17๊ฐœ ์‹ ์ฒด ๊ด€์ ˆ ์‹ค์‹œ๊ฐ„ ์ถ”์ 
โ†’ ๋‚™์ƒ ํŒจํ„ด ๊ฐ์ง€
โ†’ ๊ท ํ˜• ๋Šฅ๋ ฅ ์ •๋Ÿ‰ํ™” (Balance Score)
    โ†“
์Œ์„ฑ ๋ถ„์„ ๋ชจ๋ธ
โ†’ ์šธ์Œ์†Œ๋ฆฌ / "๋„์™€์ฃผ์„ธ์š”" / ์ถฉ๊ฒฉ์Œ ๊ฐ์ง€
    โ†“
๋‚™์ƒ ํŒ๋‹จ ๋กœ์ง
โ†’ ์˜คํƒ(False Alarm) ์ตœ์†Œํ™”
   (์•‰๋Š” ๋™์ž‘ vs ์‹ค์ œ ๋‚™์ƒ ๊ตฌ๋ณ„)
    โ†“
์•Œ๋ฆผ ์„œ๋น„์Šค (ํ‘ธ์‹œ / ๋ฌธ์ž / ์นด์นด์˜คํ†ก)
    โ†“
๊ฐ€์กฑ ์•ฑ + ๋Œ€์‹œ๋ณด๋“œ

์—ฃ์ง€ AI: ๋ชจ๋ธ์ด ์—ฃ์ง€ ๋””๋ฐ”์ด์Šค์—์„œ ์ง์ ‘ ์‹คํ–‰ โ€” ๋‚ฎ์€ ์ง€์—ฐ์‹œ๊ฐ„, ํด๋ผ์šฐ๋“œ ์—ฐ๊ฒฐ ๋ถˆํ•„์š”, ํ”„๋ผ์ด๋ฒ„์‹œ ๋ณดํ˜ธ.


๊ฒŒ์ž„: "Frozen!" โ€” ์˜ˆ๋ฐฉ์„ ๊ฒŒ์ž„์œผ๋กœ

์น˜๋ฃŒ์‹ค์—์„œ ํ•˜๋˜ ๊ท ํ˜• ํ›ˆ๋ จ์„ ์ง‘์—์„œ, ๋งค์ผ, ์žฌ๋ฏธ์žˆ๊ฒŒ.

  • ์Œ์•…์ด ๋ฉˆ์ถ”๋ฉด ๊ทธ ์ž๋ฆฌ์—์„œ ๋ฉˆ์ถ”๋Š” ๋‹จ์ˆœํ•œ ๊ฒŒ์ž„
  • YOLOv11์ด ๋ฉˆ์ถ˜ ์ž์„ธ์˜ ์•ˆ์ •์„ฑ์„ ์‹ค์‹œ๊ฐ„ ์ธก์ •
  • ์ „ ์„ธ๊ณ„ ๋ˆ„๊ตฌ๋‚˜ 5์ดˆ ๋งŒ์— ์ดํ•ด ๊ฐ€๋Šฅ โ€” ์–ธ์–ด ์žฅ๋ฒฝ ์—†์Œ
  • ๋žญํ‚น ์‹œ์Šคํ…œ์œผ๋กœ ์ง€์†์  ๋™๊ธฐ ๋ถ€์—ฌ
๋žญํ‚น ๊ตฌ์กฐ:
โ”œโ”€โ”€ ๊ฐœ์ธ ๋žญํ‚น (์ง€๋‚œ ์ฃผ ๋‚˜ vs ์ด๋ฒˆ ์ฃผ ๋‚˜)
โ”œโ”€โ”€ ์ง€์—ญ ๋žญํ‚น (์šฐ๋ฆฌ ๋™๋„ค ๋ช‡ ์œ„?)
โ””โ”€โ”€ ์—ฐ๋ น๋Œ€ ๋žญํ‚น (70๋Œ€ ์ค‘ ๋ช‡ ์œ„?)

๐Ÿ”ง ๋™์ž‘ ๋ฐฉ์‹ โ€” ๊ฒ€ํ†  ์ค‘์ธ ๋‘ ๊ฐ€์ง€ ๋ฐฉํ–ฅ

ํ˜„์žฌ ๋‘˜ ์ค‘ ํ•˜๋‚˜๋ฅผ ์„ ํƒํ•˜๊ฑฐ๋‚˜, ๊ตฌํ˜„ ๋‚œ์ด๋„์— ๋”ฐ๋ผ ๊ฒฐ์ • ์˜ˆ์ •.

Option A โ€” ์ž์œ  ๋™์ž‘ ๋ชจ๋“œ (Free Movement)

์Œ์•… ์žฌ์ƒ โ†’ ์‚ฌ์šฉ์ž๊ฐ€ ์ž์œ ๋กญ๊ฒŒ ์›€์ง์ž„
์Œ์•… ์ •์ง€ โ†’ ๊ทธ ์ž๋ฆฌ์—์„œ ๋ฉˆ์ถค
AI๊ฐ€ ๋ฉˆ์ถ˜ ์ž์„ธ์˜ ์•ˆ์ •์„ฑ(๊ท ํ˜•)๋งŒ ์ธก์ •
โ†’ ์–ด๋–ค ์ž์„ธ๋“  ์ƒ๊ด€์—†์Œ, ์–ผ๋งˆ๋‚˜ ์•ˆ ํ”๋“ค๋ฆฌ๋Š”์ง€๊ฐ€ ์ ์ˆ˜
  • ์žฅ์ : ๊ตฌํ˜„ ์‰ฌ์›€, ๋ˆ„๊ตฌ๋‚˜ ์ฆ‰์‹œ ๊ฐ€๋Šฅ, ์ง€์‹œ ๋ถˆํ•„์š”
  • ๋‹จ์ : ์žฌํ™œ ์šด๋™์œผ๋กœ์„œ์˜ ํšจ๊ณผ ํ‘œ์ค€ํ™” ์–ด๋ ค์›€

Option B โ€” ํ…œํ”Œ๋ฆฟ ๋™์ž‘ ๋ชจ๋“œ (Guided Pose)

์Œ์•… ์žฌ์ƒ โ†’ ํ™”๋ฉด์— ๋ชฉํ‘œ ์ž์„ธ(ํ…œํ”Œ๋ฆฟ) ํ‘œ์‹œ
์‚ฌ์šฉ์ž๊ฐ€ ๊ทธ ์ž์„ธ๋ฅผ ๋”ฐ๋ผ ํ•จ
์Œ์•… ์ •์ง€ โ†’ ํ•ด๋‹น ์ž์„ธ๋กœ ๋ฉˆ์ถค
AI๊ฐ€ ํ…œํ”Œ๋ฆฟ ์ž์„ธ์™€์˜ ์ผ์น˜๋„ + ์•ˆ์ •์„ฑ ๋™์‹œ ์ธก์ •
  • ์žฅ์ : ์žฌํ™œ ์šด๋™ ํšจ๊ณผ ๋ช…ํ™•, ์ž์„ธ ๊ต์ • ๊ฐ€๋Šฅ
  • ๋‹จ์ : ๊ตฌํ˜„ ๋ณต์žก (pose matching ๋กœ์ง ํ•„์š”), ์ง€์‹œ ๋”ฐ๋ฅด๊ธฐ ์–ด๋ ค์šด ์‚ฌ์šฉ์ž์—๊ฒŒ ๋ถˆํŽธ

ํ˜„์žฌ ํŒ๋‹จ:

ํ…œํ”Œ๋ฆฟ ๊ตฌํ˜„์ด ์–ด๋ ค์šธ ๊ฒฝ์šฐ โ†’ Option A (์ž์œ  ๋™์ž‘)
์—ฌ์œ ๊ฐ€ ์žˆ์„ ๊ฒฝ์šฐ โ†’ Option B (ํ…œํ”Œ๋ฆฟ) ๋˜๋Š” ๋‘ ๋ชจ๋“œ ๋ณ‘ํ–‰

๐ŸŽฏ ์‚ฌ์šฉ์ž ๋งž์ถคํ˜• ๊ฒŒ์ž„ โ€” Personalized Mode

์‚ฌ์šฉ์ž๊ฐ€ ๋ˆ„๊ตฌ์ธ์ง€ ๋จผ์ € ํŒŒ์•…ํ•˜๊ณ , ๊ทธ์— ๋งž๋Š” ๊ฒŒ์ž„ ์กฐ๊ฑด์„ ์ž๋™์œผ๋กœ ์„ค์ •ํ•œ๋‹ค.

์‚ฌ์šฉ์ž ํ”„๋กœํ•„ ์ž…๋ ฅ
โ”œโ”€โ”€ ๋‚˜์ด / ์„ฑ๋ณ„
โ”œโ”€โ”€ ํ˜„์žฌ ์ƒํƒœ ์„ ํƒ:
โ”‚   โ”œโ”€โ”€ ์ผ๋ฐ˜ ๋…ธ์ธ (๊ท ํ˜• ์œ ์ง€ ๋ชฉ์ )
โ”‚   โ”œโ”€โ”€ ๋‡Œ์กธ์ค‘ ์žฌํ™œ ํ™˜์ž
โ”‚   โ”œโ”€โ”€ ์น˜๋งค ํ™˜์ž
โ”‚   โ”œโ”€โ”€ ์ •ํ˜•์™ธ๊ณผ ์ˆ˜์ˆ  ํ›„ ํšŒ๋ณต
โ”‚   โ””โ”€โ”€ ์ง€์ฒด ์žฅ์• ์ธ (ํœ ์ฒด์–ด / ๋ณดํ–‰ ๋ณด์กฐ๊ธฐ ์‚ฌ์šฉ)
โ””โ”€โ”€ ์žฌํ™œ ๋ชฉํ‘œ ์„ค์ • (๊ท ํ˜• ์œ ์ง€ / ๋‚™์ƒ ์˜ˆ๋ฐฉ / ๊ทผ๋ ฅ ํšŒ๋ณต)

์กฐ๊ฑด๋ณ„ ๊ฒŒ์ž„ ํŒŒ๋ผ๋ฏธํ„ฐ ๋ณ€ํ™”:

์‚ฌ์šฉ์ž ์œ ํ˜• ์Œ์•… ์†๋„ ๋ฉˆ์ถค ์‹œ๊ฐ„ ์ธก์ • ๋ถ€์œ„ ํ—ˆ์šฉ ํ”๋“ค๋ฆผ
์ผ๋ฐ˜ ๋…ธ์ธ ๋ณดํ†ต 5์ดˆ ์ „์‹  ๋ณดํ†ต
๋‡Œ์กธ์ค‘ ํ™˜์ž ๋А๋ฆผ 8์ดˆ ์ƒ์ฒด ์ง‘์ค‘ ๋„“๊ฒŒ ํ—ˆ์šฉ
์น˜๋งค ํ™˜์ž ๋А๋ฆผ + ๋‹จ์ˆœ 10์ดˆ ์ „์‹  (๋‹จ์ˆœ ์ž์„ธ) ๋„“๊ฒŒ ํ—ˆ์šฉ
์ˆ˜์ˆ  ํ›„ ํšŒ๋ณต ๋งค์šฐ ๋А๋ฆผ 10์ดˆ ํ•˜์ฒด ๋ถ€๋‹ด ์ตœ์†Œํ™” ๋„“๊ฒŒ ํ—ˆ์šฉ
์ง€์ฒด ์žฅ์• ์ธ ๋ณดํ†ต 5์ดˆ ์•‰์€ ์ž์„ธ ๊ธฐ์ค€ ์ƒ์ฒด๋งŒ ์ธก์ •

์‹œ๊ฐ„์— ๋”ฐ๋ผ ์ž๋™ ๋‚œ์ด๋„ ์กฐ์ •:

1์ฃผ์ฐจ: ์‰ฌ์šด ์ž์„ธ, ๋А๋ฆฐ ์Œ์•…, ๊ธด ๋ฉˆ์ถค ์‹œ๊ฐ„
      โ†“ Balance Score ๊ฐœ์„  ๊ฐ์ง€ ์‹œ
2์ฃผ์ฐจ: ์กฐ๊ธˆ ๋” ์–ด๋ ค์šด ์ž์„ธ๋กœ ์ž๋™ ์—…๊ทธ๋ ˆ์ด๋“œ
      โ†“ ๊ฐœ์„  ์—†์„ ์‹œ
      โ†’ ๊ฐ€์กฑ์—๊ฒŒ ์•Œ๋ฆผ

3. ๊ฐœ์„ ํ–ˆ๋‹ค๋Š” ๊ฑธ ์–ด๋–ป๊ฒŒ ์ฆ๋ช…ํ•˜๋Š”๊ฐ€?

์ธก์ • ๊ฐ€๋Šฅํ•œ ์ง€ํ‘œ (Data-Driven)

โ‘  Balance Score (๊ท ํ˜• ์ ์ˆ˜)

์ธก์ • ๋ฐฉ๋ฒ•:
YOLOv11 โ†’ ์‹ ์ฒด ์ค‘์‹ฌ์ (CoM) ์ถ”์ 
โ†’ ์‹œ๊ฐ„๋‹น ํ”๋“ค๋ฆผ ๋ฒ”์œ„ (cm) ์ธก์ •
โ†’ 0~100์  ์ •๋Ÿ‰ํ™”

์ž„์ƒ ๊ธฐ์ค€: Berg Balance Scale (BBS) ๊ณผ ์ƒ๊ด€๊ด€๊ณ„ ๊ฒ€์ฆ

โ‘ก ๋‚™์ƒ ๊ฐ์ง€ ์ •ํ™•๋„

Precision / Recall / F1-Score ์ธก์ •
๋ชฉํ‘œ: Precision 90%+, Recall 95%+
(์˜คํƒ์€ ์ค„์ด๊ณ , ์‹ค์ œ ๋‚™์ƒ์€ ๋†“์น˜์ง€ ์•Š๋Š”๋‹ค)

Baseline: ๊ธฐ๋ณธ ์˜์ƒ ํ•™์Šต ๋ชจ๋ธ (standard video-based action recognition)

โ‘ข ์žฌํ™œ ์ง„ํ–‰๋„

์ฃผ๊ฐ„ Balance Score ๋ณ€ํ™”์œจ
์šด๋™ ์ง€์†๋ฅ  (๋ฉฐ์น  ์—ฐ์† ์ฐธ์—ฌํ–ˆ๋Š”๊ฐ€)
๋ณ‘์› ๋ฐฉ๋ฌธ ํšŸ์ˆ˜ ๊ฐ์†Œ์œจ

โ‘ฃ ์•Œ๋ฆผ ์‘๋‹ต ์‹œ๊ฐ„

๋‚™์ƒ ๋ฐœ์ƒ โ†’ ๊ฐ€์กฑ ์•Œ๋ฆผ ์ˆ˜์‹ ๊นŒ์ง€:
๋ชฉํ‘œ: 3์ดˆ ์ด๋‚ด

๋ฐ์ดํ„ฐ ์ˆ˜์ง‘ ์œ ํ˜•

Skeleton ๋ฐ์ดํ„ฐ:
  - 17๊ฐœ ๊ด€์ ˆ (x, y, confidence score)
  - 30 FPS ์‹ค์‹œ๊ฐ„ ์ถ”์ 

์Œ์„ฑ ๋ฐ์ดํ„ฐ:
  - MFCC features
  - 16kHz sampling rate
  - ์šธ์Œ์†Œ๋ฆฌ / ์ถฉ๊ฒฉ์Œ / "๋„์™€์ฃผ์„ธ์š”" ๋ถ„๋ฅ˜

Balance Score:
  - CoM displacement (cm/s)
  - ์„ธ์…˜ ์‹œ์ž‘ ~ ์ข…๋ฃŒ ๋ณ€ํ™”๋Ÿ‰

์ด๋ฒคํŠธ ๋กœ๊ทธ:
  - timestamp, room ID, alert type, response time

Dataset ์ „๋žต: Le2i Fall Detection Dataset + UR Fall Detection Dataset ๊ธฐ๋ฐ˜์œผ๋กœ
์ดˆ๊ธฐ ๋ชจ๋ธ ๊ฒ€์ฆ. ์ดํ›„ ์‹ค์ œ ๊ฐ€์ • ํ™˜๊ฒฝ์—์„œ ์ž์ฒด ์†Œ๊ทœ๋ชจ ๋ฐ์ดํ„ฐ ์ˆ˜์ง‘ ์˜ˆ์ •.


๋ฐ์ดํ„ฐ ์‹œ๊ฐํ™” โ€” ๋Œ€์‹œ๋ณด๋“œ

๊ฐ€์กฑ์šฉ ์•ฑ:
โ”œโ”€โ”€ ์‹ค์‹œ๊ฐ„ ์•Œ๋ฆผ (๋‚™์ƒ ๋ฐœ์ƒ ์‹œ)
โ”œโ”€โ”€ ์˜ค๋Š˜์˜ ํ™œ๋™ ์š”์•ฝ
โ””โ”€โ”€ ์ด๋ฒˆ ์ฃผ ๊ฑด๊ฐ• ํŠธ๋ Œ๋“œ ๊ทธ๋ž˜ํ”„

๋ณ‘์›/์˜์‚ฌ์šฉ:
โ”œโ”€โ”€ ์ฃผ๊ฐ„ Balance Score ๋ณ€ํ™”
โ”œโ”€โ”€ ์šด๋™ ์ฐธ์—ฌ์œจ
โ”œโ”€โ”€ ๋‚™์ƒ ๋ฐœ์ƒ ์ด๋ ฅ
โ””โ”€โ”€ AI ์ƒ์„ฑ ์žฌํ™œ ๋ฆฌํฌํŠธ (Gemini AI)

์œ„ํ—˜ ๊ตฌ์—ญ ํžˆํŠธ๋งต:
โ””โ”€โ”€ ๋‚™์ƒ ์œ„ํ—˜ Top 3 ๊ตฌ์—ญ โ€” ๊ฑฐ์‹ค, ์นจ์‹ค, ๋ณต๋„
    (๊ตฌ์—ญ๋ณ„ ๋‚™์ƒ ํŠน์„ฑ์ด ๋‹ค๋ฆ„ โ†’ ๊ตฌ์—ญ๋ณ„ ๋Œ€์‘ ๋ฐฉ์•ˆ ์ œ์•ˆ)

4. ๊ธฐ๋Œ€ ํšจ๊ณผ๋Š”?

๊ฐœ์ธ ์ฐจ์›

Before MobiCare:
๋‚™์ƒ โ†’ ์•„๋ฌด๋„ ๋ชจ๋ฆ„ โ†’ ์ˆ˜ ์‹œ๊ฐ„ ํ›„ ๋ฐœ๊ฒฌ โ†’ ๊ณจ์ ˆ โ†’ ์ž…์› โ†’ ์‚ฌํšŒ์  ๊ณ ๋ฆฝ

After MobiCare:
๊ท ํ˜• ์ €ํ•˜ ๊ฐ์ง€ โ†’ ์žฌํ™œ ์šด๋™ โ†’ ๋‚™์ƒ ์˜ˆ๋ฐฉ โ†’ ๋ชจ๋นŒ๋ฆฌํ‹ฐ ์œ ์ง€
              ๋˜๋Š”
๋‚™์ƒ ์ฆ‰์‹œ ๊ฐ์ง€ โ†’ 3์ดˆ ์•ˆ์— ๊ฐ€์กฑ์—๊ฒŒ ์•Œ๋ฆผ โ†’ ๋น ๋ฅธ ๋Œ€์‘
  • ๋‚™์ƒ ๋ฐœ์ƒ๋ฅ  ๊ฐ์†Œ (์žฌํ™œ ์šด๋™ ์ง€์† ์‹œ ์—ฐ๊ตฌ ๊ธฐ์ค€ 30~40% ๊ฐ์†Œ ๊ฐ€๋Šฅ)
  • ๋‚™์ƒ ํ›„ ๋ฐœ๊ฒฌ ์‹œ๊ฐ„: ํ‰๊ท  1์‹œ๊ฐ„ โ†’ 3์ดˆ
  • ์žฌํ™œ ์šด๋™ ์ฐธ์—ฌ์œจ: ๋ณ‘์› ๋ฐฉ๋ฌธ ๊ธฐ๋ฐ˜ ์ฃผ 1~2ํšŒ โ†’ ๋งค์ผ ๊ฐ€๋Šฅ

์‚ฌํšŒ์  ์ฐจ์›

๊ณ ๋ นํ™”์‚ฌํšŒ + ์ง€์—ญ์†Œ๋ฉธ
        โ†“
๋…๊ฑฐ ๋…ธ์ธ ์ฆ๊ฐ€
        โ†“
MobiCare
        โ†“
โ‘  ์˜๋ฃŒ๋น„ ์ ˆ๊ฐ (๋‚™์ƒ ์˜ˆ๋ฐฉ = ์ž…์› ๊ฐ์†Œ)
โ‘ก ๊ฐ€์กฑ ์‹ฌ๋ฆฌ์  ๋ถ€๋‹ด ๊ฐ์†Œ
โ‘ข ๋…ธ์ธ ์ž๋ฆฝ ์ƒํ™œ ๊ธฐ๊ฐ„ ์—ฐ์žฅ
โ‘ฃ ์š”์–‘๋ณ‘์› / ๋…ธ์ธ๋ณต์ง€์‹œ์„ค ์•ˆ์ „ ์‹œ์Šคํ…œ์œผ๋กœ ํ™•์žฅ ๊ฐ€๋Šฅ

ํ•œ๊ตญ ๊ฑด๊ฐ•๋ณดํ—˜๊ณต๋‹จ ๊ธฐ์ค€, ๋‚™์ƒ ๊ด€๋ จ ์ž…์› ์น˜๋ฃŒ๋น„๋Š”
์—ฐ๊ฐ„ ์•ฝ 1์กฐ ์› ์ด์ƒ์œผ๋กœ ์ถ”์ •๋ฉ๋‹ˆ๋‹ค.
MobiCare๋Š” ์ด ๋น„์šฉ์˜ ์ผ๋ถ€๋ฅผ ๊ตฌ์กฐ์ ์œผ๋กœ ์ค„์ผ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.


AI Mobility ๊ด€์ ์—์„œ์˜ ์˜๋ฏธ

"๋ชจ๋นŒ๋ฆฌํ‹ฐ(์ด๋™ ๋Šฅ๋ ฅ)๋Š” ๋‹จ์ˆœํžˆ ๊ฑท๋Š” ๊ฒƒ์ด ์•„๋‹™๋‹ˆ๋‹ค.
์‚ฌํšŒ์— ์ฐธ์—ฌํ•˜๊ณ , ๊ฐ€์กฑ๊ณผ ์—ฐ๊ฒฐ๋˜๊ณ , ์ธ๊ฐ„๋‹ค์šด ์‚ถ์„ ์‚ฌ๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
MobiCare๋Š” ์—ฃ์ง€ AI๋กœ ๊ทธ ๋Šฅ๋ ฅ์„ ์ง€ํ‚ต๋‹ˆ๋‹ค."

MobiCare๋Š” ๋‚™์ƒ ๊ฐ์ง€ ์•ฑ์ด ์•„๋‹™๋‹ˆ๋‹ค.
์—ฃ์ง€ AI๋กœ ๋…๊ฑฐ ๋…ธ์ธ์˜ ์‚ถ์„ ์ง€ํ‚ค๋Š” ํ”Œ๋žซํผ์ž…๋‹ˆ๋‹ค.


์š”์•ฝ

์งˆ๋ฌธ ๋‹ต๋ณ€
๋ฌด์—‡์„? ์—ฃ์ง€ AI ๋‚™์ƒ ๊ฐ์ง€ + ๊ท ํ˜• ๊ฒŒ์ž„ + ์ฆ‰๊ฐ ์•Œ๋ฆผ ์‹œ์Šคํ…œ
๋ˆ„๊ตฌ๋ฅผ ์œ„ํ•ด? ๋…๊ฑฐ ๋…ธ์ธ, ์š”์–‘๋ณ‘์›ยท๋ณต์ง€์‹œ์„ค ๊ฑฐ์ฃผ์ž
์–ธ์ œ? 24์‹œ๊ฐ„ โ€” ๋ณดํ˜ธ์ž ์—†์ด ํ˜ผ์ž ์žˆ์„ ๋•Œ ํ•ญ์ƒ
์–ด๋””์„œ? ๊ฑฐ์‹ค, ์นจ์‹ค, ๊ณ„๋‹จ, ๋ณต๋„
์™œ? ๋…๊ฑฐ ๋…ธ์ธ์˜ ๋‚™์ƒ์€ ์ˆ˜ ์‹œ๊ฐ„ ๋™์•ˆ ์•„๋ฌด๋„ ๋ชจ๋ฅธ๋‹ค
์–ด๋–ป๊ฒŒ? ๋‹จ์ผ ์นด๋ฉ”๋ผ + ๊ฒฝ๋Ÿ‰ ์—ฃ์ง€ AI + ์ฆ‰๊ฐ ์•Œ๋ฆผ (๋ฌธ์ž/ํ‘ธ์‹œ/์นด์นด์˜คํ†ก)

Team 4 ยท ์ดˆ์ฝ”ํ•˜์ž„ ยท ICCAS 2026 b60be35362b9b5af887b79fae4c329be74e55813

About

Edge AI-Based Real-Time Fall Detection and Alert System for Elderly People Living Alone

Topics

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages