VOICE RECOGNITION Video Caption
1. Product Recognition for the Blind

창의성 25점 (아이디어의 독창성 및 혁신성): 20
시장성 25점 (아이디어의 사업화 가능성 및 경제성): 20
발전성 25점 (아이디어의 지속성장 가능성): 15
완성도 25점 (아이디어 사업화/마케팅 전략, 동향/사례분석 등의 타당성 및 완성도): 20
AIM:
- In Korea, all canned beverages are labeled as "음료" or "탄산" as braille (점자) for the blind.
- There exists a clear 정보 비대칭 amongst those with disabilities and those who don't, especially since they cannot choose their drinks by their preference without someone's help
- This app aims to recognize products (not limited to canned goods) on shelves with the user's phone-camera.
- The app will contain a real-time image recognition algorithm that once focused on a product, the app will identify what the product is (maybe other info as well - prices etc.).
2. 수능 맞춤형 문제 - 이과 과목 위주 (수학, 과학)
창의성 25점 (아이디어의 독창성 및 혁신성): 15
시장성 25점 (아이디어의 사업화 가능성 및 경제성): 20
발전성 25점 (아이디어의 지속성장 가능성): 18
완성도 25점 (아이디어 사업화/마케팅 전략, 동향/사례분석 등의 타당성 및 완성도): 20
AIM:
- 
This program first provides 2 practice-exams that test the user's familiarity with certain types of questions 
- 
With the student's scores in the 2-3 exams provided at the start of the program, the program tracks the total correct answers (by percentage) within its respective question type 
- 
Through a machine learning algorithm, the program assigns weights to each category depending on its respective user-score. ⇒ The program learns the skill-sets of the users and gives a practice-exam that has more number of questions on those categories on which the user got wrong 
POSSIBLE ROUTE FOR COMPETITIVENESS:
- The program would be made online on which many other users can compete with each others' progress
- the progress score will be measured by (total # correct so far / total # solved) per category
 
PROBLEMS:
- Copy-right issues with problem sets provided by test-taking institutes (수능은 이 문제 없을듯?)