Skip to content
/ witt Public

Real-time object detection services through Android mobile camera. plus, ChatGPT, Translator, OCR, and STT services are supported with YOLO v5 & YOLO v8

Notifications You must be signed in to change notification settings

CAVASOL/witt

Repository files navigation

WITT

WITT는 기발함을 의미하는 단어 "Wit" 와 "What is that" 의 합성어로, 딥러닝 기술을 활용하여 다양한 사용자들에게 편의를 제공하는 서비스를 구현하자는 취지를 담고 있습니다. WITT의 핵심 기능은 안드로이드 모바일 카메라를 통한 실시간 객체 탐지 및 이미지 탐색으로서, 사용자가 자율적으로 주변 환경과 객체를 탐색할 수 있도록 서비스를 지원합니다. 더불어, 탐색한 객체에 대한 정보를 제공받을 수 있도록 GPT, 번역(한국어, 영어, 일본어, 중국어), OCR(Optical Character Recognition) 및 STT(Speech To Text) 서비스를 지원합니다.

WITT, a portmanteau of Wit representing ingenuity and What is that, aims to implement a service utilizing deep learning technology to provide convenience to diverse users. Its core functionality lies in real-time object detection and image search via Android mobile camera , enabling users to autonomously explore their surroundings and objects. Additionally, it supports services such as GPT for obtaining information about the detected objects, translation (Korean, English, Japanese, Chinese), OCR (Optical Character Recognition), and STT (Speech To Text)."

Why you need WITT

  • 혁신적 기술과 시각화를 통한 교육적 접근성 향상
    Improving educational accessibility through innovative technology and visualization

  • 사전 학습이 필요없는 환경 및 물체 탐색
    Explore environments and objects without prior learning

  • 개별화된 교육환경으로의 변화
    Change to an individualized educational environment

WITT Goal

  • 바운딩 박스를 통한 실시간 객체 탐지 서비스 지원
    Supports real-time object detection service through bounding box

  • 사용자 주도적 환경 탐색 및 학습 경험 제공
    Provides user-driven environmental exploration and learning experience

  • 다차원적 도구로서의 활용
    Use as a multidimensional tool

Artifact

WITT 프로젝트 아티팩트 폴더는 본 리포지토리에 포함되어 있으며, 주요 디렉토리 구성은 다음과 같습니다.
상세한 아티팩트 디렉토리 구성 및 실험 결과와 관련한 정보는 링크를 클릭하여 확인해 주세요.

artifact
    ├── Yolov5l
    ├── Yolov8n
    ├── animals_Y5l.ipynb
    ├── animals_Y8n.ipynb
    ├── Augmentation.ipynb
    └── README.md

Development environment

  • IDE: VS Code
  • Version Control: Git & Github
  • Models: YOLO v5l, YOLO v8n
  • Technology Stack: Flutter Framework, Dart, Flutter Packages, Firebase, HTTP

Demo

WITT application screen composition and functions gif

WITT Screen Composition and Functions

Welcome Sign-in Home
WITT application welcome screen image WITT application sign-in screen image WITT application home screen image
Home Explore Image Detection (YOLO v5l)
WITT application home screen image WITT application explore screen image WITT application object detection results as label screen image
Camera Live Detection (YOLO v8n) Image Detection (YOLO v8n)
WITT application camera screen image WITT application camera live detetion screen image WITT application image object detection screen image
Chat GPT GPT with Korean GPT with English
WITT application chat gpt screen image WITT application gpt with Korean screen image WITT application gpt with English screen image
Translator 캡쳐하기 (OCR) 캡쳐한 후 번역 (OCR)
WITT application translator screen image WITT application optical character recognition image capture screen image WITT application optical character recognition results screen image

About

Real-time object detection services through Android mobile camera. plus, ChatGPT, Translator, OCR, and STT services are supported with YOLO v5 & YOLO v8

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published