Deep Learning

Image Classification

This deep learning model uses a pre-trained InceptionV3 network with additional custom layers for transfer learning. The model achieved 94% accuracy on the test dataset.

Model Accuracy

94%

Model Type

CNN + Transfer Learning

Category

Deep Learning

Tech Stack

4 Tools

Key Features

  • Transfer learning with InceptionV3
  • Data augmentation techniques
  • Custom CNN architecture
  • Model ensemble for improved accuracy
  • Grad-CAM visualization for interpretability

Technologies

TensorFlowKerasOpenCVNumPy

Performance Metrics

Accuracy

94

Precision

93

Recall

94

F1-Score

93.5

Training History

Class Distribution

Interested in this project?