How gokulnathsuresh05/senior_citizen_identification works
This repository is a face-based senior-citizen and gender classifier, with both a GUI and a command-line image workflow. Its other major product branch prepares UTKFace data, trains the two-output CNN, and evaluates it. The README documents the high-level detection pipeline and training steps; source excerpts confirm several implementation details, but do not show all GUI and training internals.
Main entry points: detect_and_predict.py Image Detection CLI · train_model.py Training Runner · gui.py Desktop GUI · senior_gender_model.keras Trained Model
Components 20
Image Inference
- Desktop GUI
app/gui.py - Image Detection CLI
src/detect_and_predict.py - Face Detection
src/detect_and_predict.py - Detection Filtering
src/detect_and_predict.py - Duplicate Removal
src/detect_and_predict.py - Face Preprocessing
src/detect_and_predict.py - Annotated Result
Model Training
- CNN Classifier
src/build_model.py - Trained Model
models/senior_gender_model.keras - Training Runner
src/train_model.py - Data Pipeline
src/data_generator.py
Dataset Preparation
- Face Dataset Creator
src/create_detected_dataset.py - Label Analysis
src/analyze_labels.py - Dataset Check
src/check_dataset.py - Dataset Split
src/split_dataset.py - Class Weights
src/calculate_class_weights.py
Model Evaluation
- Model Evaluation
src/evaluate_model.py
Other components
- User
- Input Image
- Test Dataset
Connections 25
- User → Desktop GUI: opens
README.md - User → Image Detection CLI: runs
README.md - Input Image → Desktop GUI: selects
app/gui.py - Input Image → Image Detection CLI: processes
README.md - Image Detection CLI → Face Detection: detects faces
README.md - Face Detection → Detection Filtering: filters
README.md - Detection Filtering → Duplicate Removal: removes duplicates
README.md - Duplicate Removal → Face Preprocessing: crops and preprocesses
README.md - Face Preprocessing → CNN Classifier: predicts
README.md - Desktop GUI → Trained Model: loads model
- Image Detection CLI → Trained Model: loads model
- CNN Classifier → Annotated Result: generates annotations
README.md - Desktop GUI → Annotated Result: saves image
app/gui.py - Image Detection CLI → Annotated Result: saves image
README.md - User → Face Dataset Creator: prepares faces
- User → Label Analysis: analyzes labels
README.md - User → Dataset Check: checks data
README.md - User → Dataset Split: splits data
README.md - User → Class Weights: calculates weights
README.md - Dataset Split → Training Runner: provides splits
README.md - Training Runner → Data Pipeline: uses batches
- Training Runner → CNN Classifier: builds model
src/train_model.py - Training Runner → Trained Model: saves model
README.md - Test Dataset → Model Evaluation: supplies samples
src/evaluate_model.py - Model Evaluation → Trained Model: loads model