How hibatillah/deep-learning works
Main Components
A. Frontend Layer (Next.js):
• Client-side Components (/web/app/_components/)
• Audio Classification Component
• Image Classification Component
• Text Sentiment Component
• UI Components Layer (/web/components/ui/)
• Theme Management (/web/components/theme/)
• Custom Components (/web/components/custom/)
B. Backend Layer (Flask):
• Flask API Server (app.py)
• Deep Learning Models:
• Gun Audio Model
• IMDB Sentiment Model
• Landscape Images Model
• [Client Tier] → [API Tier] → [Model Tier]
Specific Elements
Frontend Section:
• Three main classification components
• UI component library section
• Theme management system
API Layer:
• Flask server
• API endpoints for each prediction type
• Request/response flow
Model Layer:
• Model file locations
• Model input/output types
Connection Types
• HTTP requests between frontend and API
• Model inference calls
• Data flow directions
• Component dependencies
Additional Details
Technical specifications:
• Next.js v14.2.17
• Flask v3.1.0
• Model file formats (.h5)
• Port numbers (3000 for web, 8000 for API)