Image-Classification-using-Convolutional-Neural-Networks #815
Conversation
|
Our team will soon review your PR. Thanks @UTSAVS26 :) |
abhisheks008
added
Status: Requested Changes
…ing-Convolutional-Neural-Networks
|
Changes done successfully |
|
Follow the README template and update the README as per the given template. |
…ing-Convolutional-Neural-Networks
Changes done for the README.md file based on template |
|
Your model implementations are good enough, but the accuracy scores a bit concerning especially for the CIFAR dataset. Can come up with a model, which is having at least 80% accuracy with the CIFAR dataset? |
I will try to get some new models or try to make one if necessary for CIFAR10 dataset but for now i have get these accuracies as reported above and i have added separate precision and classification report for each of the model individual too. |
Pull Request for DL-Simplified 💡
Issue #730
Issue Title: Image Classification using Convolutional Neural Networks
Info about the related issue (Aim of the project): The goal of this project is to implement and compare the performance of various deep learning models, including LeNet-5, MobileNet, ResNet50, Simple CNN, and VGG16, for image classification tasks. By training these models on datasets like CIFAR-10 and MNIST, the project aims to evaluate and analyze their respective accuracies, loss values, and generalization capabilities. This comparative study will highlight the strengths and weaknesses of each architecture and provide insights into their effectiveness for practical image classification applications. The findings will guide the selection of the most suitable model architecture for robust image classification.
Name: Utsav Singhal
GitHub ID: UTSAVS26
Email ID: [email protected]
Identify yourself: SSoC 2024 participant
Closes: #730
Describe the add-ons or changes you've made 📃
This project involves the implementation and comparison of five deep learning models: LeNet-5, MobileNet, ResNet50, Simple CNN, and VGG16. Each model is designed for image classification tasks and has been tested on datasets like CIFAR-10 and MNIST. The models were trained with data augmentation to improve generalization and their performances were evaluated based on metrics such as accuracy, precision, recall, and F1-score. The project aims to identify which model best balances accuracy and generalization, offering insights into the advantages of different architectural features like residual connections and depthwise separable convolutions in deep learning.
Type of change ☑
What sort of change have you made:
How Has This Been Tested? ⚙
The project involved evaluating five deep learning models—LeNet-5, MobileNet, ResNet50, Simple CNN, and VGG16—for image classification tasks. The models were trained on the CIFAR-10 and MNIST datasets, with the data split into training, validation, and test sets. Training data underwent extensive augmentation to enhance model generalization. Each model was trained using categorical cross-entropy loss and the Adam optimizer, with callbacks for early stopping, best model checkpointing, and learning rate reduction. Performance was assessed on test sets using accuracy, precision, recall, F1-score, and confusion matrices. Results were visualized through accuracy curves and detailed classification reports. This comprehensive evaluation aimed to identify the most effective architecture for robust image classification.
Checklist: ☑