https://github.com/elmezianech/pizza_steak_imageclassification_cnn

This project introduces a powerful image classification model to distinguish between pizza and steak images. Leveraging advanced techniques, the model achieves robust performance in handling complex visual features. With an accuracy of 87.40% on the validation dataset.
https://github.com/elmezianech/pizza_steak_imageclassification_cnn

cnn data-augmentation image-classification keras tensorflow

Last synced: 14 days ago
JSON representation

This project introduces a powerful image classification model to distinguish between pizza and steak images. Leveraging advanced techniques, the model achieves robust performance in handling complex visual features. With an accuracy of 87.40% on the validation dataset.

• Host: GitHub
• URL: https://github.com/elmezianech/pizza_steak_imageclassification_cnn
• Owner: elmezianech
• Created: 2024-01-21T21:00:09.000Z (about 1 year ago)
• Default Branch: main
• Last Pushed: 2024-01-21T21:21:53.000Z (about 1 year ago)
• Last Synced: 2024-11-15T23:26:56.269Z (3 months ago)
• Topics: cnn, data-augmentation, image-classification, keras, tensorflow
• Language: Jupyter Notebook
• Homepage:
• Size: 585 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Pizza_Steak_ImageClassification_CNN

This project introduces a powerful image classification model to distinguish between pizza and steak images. Leveraging advanced techniques, the model achieves robust performance in handling complex visual features.

Link of Data : https://www.kaggle.com/datasets/kelixirr/pizza-steak-image-classification-dataset/data

Key Techniques Employed

- Convolutional Neural Network (CNN)

Implemented a sophisticated CNN architecture utilizing TensorFlow and Keras, designed to learn intricate hierarchical patterns in the images.

- Data Augmentation

Applied data augmentation techniques to enrich the training dataset, enhancing the model's ability to generalize to various image variations.

- Dropout Layer

Incorporated dropout layers to combat overfitting, ensuring the model's robustness by preventing reliance on specific features.

- Model Evaluation

Evaluated model performance in different scenarios:

Model 1: Created a baseline CNN model.

Model 2: Implemented data augmentation and added a dropout layer to observe improvements.

Results 

- Model 1

Epoch 5:

Training Loss: 0.2933 | Accuracy: 88.40%

Validation Loss: 0.3474 | Accuracy: 85.40%

- Model 2 (with Data Augmentation and Dropout)

Epoch 5:

Training Loss: 0.4697 | Accuracy: 78.80%

Validation Loss: 0.3307 | Accuracy: 87.40%

These results showcase the progression and improvements observed in Model 2 after incorporating data augmentation and dropout layers. Model 2 demonstrates enhanced generalization and reduced overfitting, as reflected in the validation accuracy.