{"id":44393997,"url":"https://github.com/rashad-malik/cifar-10-image-classification","last_synced_at":"2026-02-12T03:00:56.073Z","repository":{"id":319839431,"uuid":"1079782262","full_name":"rashad-malik/CIFAR-10-Image-Classification","owner":"rashad-malik","description":"A deep learning project implementing a custom neural network architecture for CIFAR-10 image classification, achieving 86.84% test accuracy through iterative improvements and optimisation techniques.","archived":false,"fork":false,"pushed_at":"2025-10-20T11:49:01.000Z","size":263,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2025-10-20T13:28:48.708Z","etag":null,"topics":["cifar10","computer-vision","deep-learning","image-classification","machine-learning","pytorch"],"latest_commit_sha":null,"homepage":"https://rashad-malik.github.io/CIFAR-10-Image-Classification/","language":"HTML","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/rashad-malik.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2025-10-20T11:32:39.000Z","updated_at":"2025-10-20T11:49:05.000Z","dependencies_parsed_at":"2025-10-20T13:28:58.452Z","dependency_job_id":null,"html_url":"https://github.com/rashad-malik/CIFAR-10-Image-Classification","commit_stats":null,"previous_names":["rashad-malik/cifar-10-image-classification"],"tags_count":null,"template":false,"template_full_name":null,"purl":"pkg:github/rashad-malik/CIFAR-10-Image-Classification","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/rashad-malik%2FCIFAR-10-Image-Classification","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/rashad-malik%2FCIFAR-10-Image-Classification/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/rashad-malik%2FCIFAR-10-Image-Classification/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/rashad-malik%2FCIFAR-10-Image-Classification/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/rashad-malik","download_url":"https://codeload.github.com/rashad-malik/CIFAR-10-Image-Classification/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/rashad-malik%2FCIFAR-10-Image-Classification/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":29355825,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-02-12T01:03:07.613Z","status":"online","status_checked_at":"2026-02-12T02:00:06.911Z","response_time":55,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["cifar10","computer-vision","deep-learning","image-classification","machine-learning","pytorch"],"created_at":"2026-02-12T03:00:21.341Z","updated_at":"2026-02-12T03:00:56.060Z","avatar_url":"https://github.com/rashad-malik.png","language":"HTML","funding_links":[],"categories":[],"sub_categories":[],"readme":"# CIFAR-10 Image Classification with Custom Neural Network Architecture\n\nA deep learning project implementing a custom neural network architecture for CIFAR-10 image classification, achieving **86.84% test accuracy** through iterative improvements and optimisation techniques.\n\n## 📋 Project Overview\n\nThis project demonstrates the development and optimisation of a neural network architecture called **RashadNet** for classifying images in the CIFAR-10 dataset. The architecture features a design with parallel convolutional paths and dynamic feature weighting, combined with modern training techniques to achieve strong performance.\n\n## 🎯 Results\n\n| Model Version | Key Features | Test Accuracy |\n|--------------|--------------|---------------|\n| Baseline | 3 blocks, basic architecture | 42.10% |\n| First Wave | Data augmentation, batch norm, dropout, 6 blocks | 56.14% |\n| **Final Model** | **Label smoothing, cosine annealing, MaxPool, 10 blocks** | **86.84%** |\n\n## 🔧 Technologies Used\n\n- **Python 3.11**\n- **PyTorch**: Deep learning framework\n- **torchvision**: Dataset loading and image transformations\n- **matplotlib**: Visualisation\n- **CUDA**: GPU acceleration\n\n## 📊 Dataset\n\nThe [CIFAR-10 dataset](https://www.cs.toronto.edu/~kriz/cifar.html) contains:\n- 60,000 32×32 colour images\n- 10 classes: airplane, automobile, bird, cat, deer, dog, frog, horse, ship, truck\n- 50,000 training images\n- 10,000 test images\n\n## 🚀 Key Features \u0026 Techniques\n\n### Data Augmentation\n- Random horizontal flips\n- Random cropping with padding\n- Normalisation to [-1, 1] range\n\n### Regularisation\n- Batch normalisation after convolutions\n- Dropout (p=0.3) in output block\n- Label smoothing (0.1)\n\n### Training Optimisation\n- Adam optimiser\n- Cosine annealing learning rate scheduler\n- Cross-entropy loss with label smoothing\n\n### Architecture Enhancements\n- Deeper network (10 intermediate blocks)\n- Wider network (64 base channels)\n- MaxPool downsampling for feature preservation\n\n## 📁 Project Structure\n\n```\nCIFAR-10 Image Classification/\n├── html_export/\n│   └── CIFAR10_image_classification.html     # Notebook exported as HTML\n├── notebook/\n│   └── CIFAR10_image_classification.ipynb    # Main Jupyter notebook\n└── README.md                                   # Project documentation\n```\n\n## 🔬 Methodology\n\nThe project follows a structured approach:\n\n1. **Dataset Preparation**: Loading and preprocessing CIFAR-10 with PyTorch DataLoaders\n2. **Basic Architecture**: Implementing the custom RashadNet with intermediate blocks\n3. **Training \u0026 Testing**: Establishing baseline performance and evaluation metrics\n4. **Iterative Improvements**: Systematic enhancements through two major update waves\n\n## 🙏 Acknowledgements\n\n- CIFAR-10 dataset creators\n- PyTorch community and documentation\n- Inspiration from modern CNN architectures (ResNet, DenseNet, etc.)","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Frashad-malik%2Fcifar-10-image-classification","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Frashad-malik%2Fcifar-10-image-classification","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Frashad-malik%2Fcifar-10-image-classification/lists"}