How rojasantino/bank_document_classifier works
This repository is a bank-document classification and cheque-field extraction system. Its main workflow prepares labeled document images, trains and evaluates classifiers, then serves uploaded documents through a REST API or Streamlit dashboard. Classification can be supplemented with Grad-CAM explanations and OCR extraction. The graph includes documented capabilities whose implementation was not fully sampled; edges are limited to README statements and source-index references, so some internal runtime details remain unspecified.
Main entry points: config.py Project configuration · main.py REST API · gradcam.py Grad-CAM explanation · dataset.py Dataset and splits
Components 21
Data preparation
- Raw documents
data/raw - Synthetic generator
data/generate_synthetic_data.py - Image preprocessing
src/data_preprocessing.py - Processed documents
data/processed - Dataset and splits
src/dataset.py
Model lifecycle
- Model training
src/train.py - CNN architectures
src/models/transfer_models.py - Custom CNN
src/models/custom_cnn.py - Model evaluation
src/evaluate.py
Inference features
- Grad-CAM explanation
src/gradcam.py - Cheque OCR extraction
src/ocr_extraction.py
User interfaces
- REST API
api/main.py - Response schemas
api/schemas.py - Streamlit dashboard
frontend/streamlit_app.py - API web page
frontend/index.html
Configuration and artifacts
- Model checkpoints
outputs/models - Evaluation reports
outputs/reports - Grad-CAM images
outputs/gradcam - Project configuration
config.py
Other components
- Document user
- Tesseract OCR
Connections 38
- Document user → REST API: uploads document
README.md - Document user → Streamlit dashboard: selects document
README.md - Synthetic generator → Raw documents: generates images
README.md - Raw documents → Image preprocessing: preprocesses
README.md - Image preprocessing → Processed documents: writes images
README.md - Processed documents → Dataset and splits: loads preferred data
src/dataset.py - Raw documents → Dataset and splits: loads fallback data
src/dataset.py - Dataset and splits → Model training: provides batches
src/train.py - Model training → CNN architectures: builds models
src/train.py - CNN architectures → Custom CNN: constructs baseline
src/models/transfer_models.py - Model training → Model checkpoints: saves checkpoints
README.md - Dataset and splits → Model evaluation: provides test split
src/evaluate.py - Model evaluation → CNN architectures: builds models
src/evaluate.py - Model evaluation → Model checkpoints: loads checkpoints
src/evaluate.py - Model evaluation → Evaluation reports: writes metrics
README.md - REST API → CNN architectures: loads classifier
api/main.py - REST API → Grad-CAM explanation: requests explanation
api/main.py - REST API → Cheque OCR extraction: requests extraction
api/main.py - REST API → Response schemas: formats responses
api/main.py - REST API → API web page: serves page
api/main.py - Streamlit dashboard → CNN architectures: loads classifier
frontend/streamlit_app.py - Streamlit dashboard → Grad-CAM explanation: runs explanation
frontend/streamlit_app.py - Streamlit dashboard → Cheque OCR extraction: offers extraction
frontend/streamlit_app.py - Grad-CAM explanation → CNN architectures: builds model
src/gradcam.py - Grad-CAM explanation → Dataset and splits: uses evaluation transforms
src/gradcam.py - Grad-CAM explanation → Model checkpoints: loads checkpoint
README.md - Grad-CAM explanation → Grad-CAM images: writes heatmap
README.md - Cheque OCR extraction → Image preprocessing: uses image operations
src/ocr_extraction.py - Cheque OCR extraction → Tesseract OCR: recognizes text
README.md - Synthetic generator → Project configuration: reads settings
data/generate_synthetic_data.py - Image preprocessing → Project configuration: reads paths
src/data_preprocessing.py - Dataset and splits → Project configuration: reads classes and splits
src/dataset.py - Model training → Project configuration: reads settings
src/train.py - Model evaluation → Project configuration: reads settings
src/evaluate.py - REST API → Project configuration: reads paths and classes
api/main.py - Streamlit dashboard → Project configuration: reads paths and classes
frontend/streamlit_app.py - Grad-CAM explanation → Project configuration: reads paths and classes
src/gradcam.py - Cheque OCR extraction → Project configuration: reads settings
src/ocr_extraction.py