{"id":26999837,"url":"https://github.com/tristan296/deepplate","last_synced_at":"2025-04-04T03:17:59.452Z","repository":{"id":285721482,"uuid":"959094086","full_name":"Tristan296/DeepPlate","owner":"Tristan296","description":"A custom-trained YOLOv11-based license plate detection with PaddleOCR for accurate ANPR.","archived":false,"fork":false,"pushed_at":"2025-04-02T09:52:14.000Z","size":18633,"stargazers_count":1,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-04-02T10:35:14.856Z","etag":null,"topics":["cv2","numpy","paddleocr","sqlite3","ultralytics","yolov11"],"latest_commit_sha":null,"homepage":"","language":"Python","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/Tristan296.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}},"created_at":"2025-04-02T08:59:50.000Z","updated_at":"2025-04-02T09:52:17.000Z","dependencies_parsed_at":"2025-04-02T10:45:41.880Z","dependency_job_id":null,"html_url":"https://github.com/Tristan296/DeepPlate","commit_stats":null,"previous_names":["tristan296/deepplate"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Tristan296%2FDeepPlate","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Tristan296%2FDeepPlate/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Tristan296%2FDeepPlate/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Tristan296%2FDeepPlate/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Tristan296","download_url":"https://codeload.github.com/Tristan296/DeepPlate/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":247112756,"owners_count":20885606,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","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":["cv2","numpy","paddleocr","sqlite3","ultralytics","yolov11"],"created_at":"2025-04-04T03:17:58.797Z","updated_at":"2025-04-04T03:17:59.440Z","avatar_url":"https://github.com/Tristan296.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Image Detection\n\n\u003e A project that utilizes machine learning to detect and classify objects in images.\n\u003e Live demo [_here_](https://www.example.com). \n\n## Table of Contents\n\n* [General Info](#general-information)\n* [Technologies Used](#technologies-used)\n* [Features](#features)\n* [Screenshots](#screenshots)\n* [Setup](#setup)\n* [Usage](#usage)\n* [Project Status](#project-status)\n* [Model Training Summary](#model-training-summary)\n* [Room for Improvement](#room-for-improvement)\n* [Acknowledgements](#acknowledgements)\n* [Contact](#contact)\n\n## General Information\n\n- This project aims to simplify the process of detecting and classifying objects in images using advanced machine learning techniques.\n- It solves the problem of manual image classification by automating the process.\n- The purpose of this project is to provide an easy-to-use tool for developers and researchers working with image data.\n\n## Technologies Used\n\n- Python - version 3.9\n- TensorFlow - version 2.10\n- OpenCV - version 4.5\n\n## Features\n\nList the ready features here:\n\n- Object detection in real-time\n- Classification of detected objects\n- Support for multiple image formats\n\n## Screenshots\n\n![Example screenshot](./img/screenshot.png)\n\n\u003c!-- Add actual screenshots of the project --\u003e\n\n## Setup\n\n- Install the required dependencies listed in `requirements.txt`.\n- Clone the repository and navigate to the project directory.\n- Run the following command to install dependencies:\n  ```\n  pip install -r requirements.txt\n  ```\n\n## Usage\n\n- To detect objects in an image, run the following command:\n  ```\n  python detect.py --image path/to/image.jpg\n  ```\n- For real-time detection using a webcam:\n  ```\n  python detect.py --webcam\n  ```\n\n## Project Status\n\nProject is: _in progress_. Further improvements and optimizations are being worked on.\n\n## Room for Improvement\n\nRoom for improvement:\n\n- Enhance the accuracy of object detection.\n- Add support for additional machine learning models.\n\nTo do:\n\n- Implement a user-friendly GUI.\n- Add functionality for batch image processing.\n\n# Model Training Summary\n\n## General Information\nThe model was trained using the YOLO framework with the following configuration:\n- **Task**: Object Detection\n- **Model**: YOLO (yolo11n.pt)\n- **Dataset**: COCO8\n- **Epochs**: 100\n- **Batch Size**: 16\n- **Image Size**: 640x640\n- **Device**: MPS (Metal Performance Shaders)\n\n## Training Results\nThe training process generated the following key results:\n- **Final Model**: Saved as `last.pt` in the `weights/` directory.\n- **Metrics**:\n    - **Precision**: Improved steadily, reaching a maximum of ~0.00367.\n    - **Recall**: Peaked at 0.66667 during early epochs.\n    - **mAP@50**: Reached a maximum of 0.28376.\n    - **mAP@50-95**: Reached a maximum of 0.13616.\n\n## Loss Trends\n- **Box Loss**: Decreased significantly over epochs, indicating better localization.\n- **Classification Loss**: Reduced steadily, showing improved classification accuracy.\n- **DFL Loss**: Gradually decreased, reflecting better distribution-focused learning.\n\n## Visualizations\n- **Confusion Matrices**: `confusion_matrix.png` and `confusion_matrix_normalized.png` provide insights into class-wise performance.\n- **Curves**:\n    - `F1_curve.png`: Shows the F1-score progression.\n    - `P_curve.png`: Precision curve.\n    - `R_curve.png`: Recall curve.\n    - `PR_curve.png`: Precision-Recall curve.\n\n## Training Batches\n- Sample training images are available (`train_batch0.jpg`, `train_batch1.jpg`, etc.), showing the model's predictions during training.\n\n## Validation Results\n- Validation images (`val_batch0_labels.jpg`, `val_batch0_pred.jpg`) demonstrate the model's performance on unseen data.\n\n## Observations\n- The model shows steady improvement in metrics over epochs.\n- The mAP values suggest room for improvement in detection accuracy.\n- The training process was well-documented with visualizations and logs.\n\n## Next Steps\n- Fine-tune the model to improve mAP and precision.\n- Experiment with different hyperparameters or augmentations.\n- Test the model on real-world data to evaluate its generalization.\n\n\n## Acknowledgements\n\n### YOLO by Ultralytics\n\n```bibtex\n@misc{yolo2023,\n  author       = {Jocher, Glenn and Chaurasia, Ayush and Qiu, Jing},\n  title        = {YOLO by Ultralytics},\n  year         = {2023},\n  publisher    = {GitHub},\n  journal      = {GitHub repository},\n  howpublished = {\\url{https://github.com/ultralytics/yolov5}}\n}\n```\n\n### PaddleOCR\n\n```bibtex\n@article{paddleocr2021,\n  author       = {PaddleOCR Contributors},\n  title        = {PaddleOCR: An Open-Source Optical Character Recognition Tool Based on PaddlePaddle},\n  year         = {2021},\n  journal      = {GitHub repository},\n  howpublished = {\\url{https://github.com/PaddlePaddle/PaddleOCR}}\n}\n```\n\n## Contact\n\nCreated by [@tristan](https://github.com/tristan296) - feel free to contact me!\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ftristan296%2Fdeepplate","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Ftristan296%2Fdeepplate","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ftristan296%2Fdeepplate/lists"}