How sajid73/avisight-bird_species_classifier works
This is a bird-species image-classification project: it combines a Streamlit prediction interface with model-training experiments and dataset/evaluation utilities. The principal user workflow is image upload, preprocessing, model prediction, and display of a species label. The README describes notebook training and evaluation capabilities, but those notebook contents were not sampled, so their internal wiring is not shown; the referenced model artifacts and upload directory are also absent from the tree.
Main entry points: app.py Streamlit app · Models Training experiments · All_folders.py Folder preparation · model_testing.ipynb Model testing
Components 12
Image inference
- Streamlit app
app.py - Image preprocessing
app.py - Model prediction
app.py - Species labels
app.py - Uploaded images
- VGG16 artifact
Model experiments
- Training experiments
Models
Dataset and evaluation
- Folder preparation
Utils/All_folders.py - Folder names
Utils/folders_name.txt - Classification utilities
Utils/Bird_Classification.py - Model testing
Utils/model_testing.ipynb
Other components
- Image user
Connections 10
- Image user → Streamlit app: uploads image
- Streamlit app → Uploaded images: writes image
- Streamlit app → Image preprocessing: calls
- Image preprocessing → Model prediction: provides image
- Model prediction → VGG16 artifact: predicts with
- Model prediction → Species labels: maps class
- Streamlit app → Image user: shows result
- Training experiments → VGG16 artifact: saves model
- Folder preparation → Folder names: writes names
- Model testing → VGG16 artifact: loads model