How officialnavneetdas/vision-based-books-recommendation-system--vbbrs- works
This Streamlit application detects books in either a live webcam stream or an uploaded image. The user selects an input mode and confidence threshold; webcam frames or uploaded images are passed to the cached YOLO detector, then annotated and shown with a book count. The ONNX weights are loaded from the repository’s model file. The source excerpt shows both detection paths and their rendering; no README or other product modules were provided.
Main entry points: app.py Book Detector · best.onnx ONNX Weights
Components 10
User Interface
- Mode Controls
app.py - Live Annotation
app.py - Image Results
app.py
Detection Runtime
- Frame Processor
app.py - Model Loader
app.py - ONNX Weights
models/best.onnx - Book Detector
app.py
Input Sources
- Webcam Session
app.py - Image Upload
app.py
Other components
- User
Connections 10
- User → Mode Controls: selects mode
app.py - Mode Controls → Webcam Session: starts session
app.py - Mode Controls → Image Upload: opens upload
app.py - Model Loader → ONNX Weights: loads weights
app.py - Webcam Session → Frame Processor: dispatches frames
app.py - Frame Processor → Book Detector: runs inference
app.py - Image Upload → Book Detector: runs inference
app.py - Model Loader → Book Detector: creates model
app.py - Frame Processor → Live Annotation: returns annotated frame
app.py - Book Detector → Image Results: plots detections
app.py