From 5f148ea27bb2940a785d3e9e46626aa7b9bb02c2 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Iv=C3=A1n=20de=20Paz?= Date: Tue, 8 Oct 2024 18:07:21 +0200 Subject: [PATCH] Changed link to DOI and added references to docs. --- README.md | 37 +++++++++++++++++++++---- docs/references.md | 69 +++++++++++++++++++++++++++------------------- 2 files changed, 72 insertions(+), 34 deletions(-) diff --git a/README.md b/README.md index e52a19f..b5a5814 100644 --- a/README.md +++ b/README.md @@ -5,7 +5,8 @@ ![Test Status](https://github.com/ipazc/mtcnn/actions/workflows/tests.yml/badge.svg) ![Pylint Check](https://github.com/ipazc/mtcnn/actions/workflows/pylint.yml/badge.svg) ![PyPI Downloads](https://img.shields.io/pypi/dm/mtcnn) -![DOI](https://zenodo.org/badge/DOI/10.5281/zenodo.13901379.svg) +[![DOI](https://zenodo.org/badge/DOI/10.5281/zenodo.13901378.svg)](https://doi.org/10.5281/zenodo.13901378) + ## Overview @@ -84,7 +85,22 @@ The full documentation for this project is available at [Read the Docs](http://m ## Citation -If you use this library for your research or projects, please consider citing the original work: +If you use this library implementation for your research or projects, please consider using this cite: + +``` +@software{ivan_de_paz_centeno_2024_13901378, + author = {Iván de Paz Centeno}, + title = {ipazc/mtcnn: v1.0.0}, + month = oct, + year = 2024, + publisher = {Zenodo}, + version = {v1.0.0}, + doi = {10.5281/zenodo.13901378}, + url = {https://doi.org/10.5281/zenodo.13901378} +} +``` + +And the original research work from Kaipeng Zhang: ``` @article{7553523, @@ -108,15 +124,24 @@ You may also reference the original GitHub repository that this project was base And the FaceNet's implementation that served as inspiration: [Facenet's MTCNN implementation](https://github.com/davidsandberg/facenet/tree/master/src/align) -## About this project -The code for this project was created to standardize face detection and provide an easy-to-use framework that helps the research community push the boundaries of AI knowledge. Learn more about the author of this code on [Iván de Paz Centeno's website](https://ipazc.com) +## About the Author -If you find this project useful, please consider supporting it through GitHub Sponsors. +This project is developed and maintained by [Iván de Paz Centeno](https://ipazc.com), with the goal of standardizing face detection and providing an easy-to-use framework to help the research community push the boundaries of AI knowledge. + +If you find this project useful, please consider supporting it through GitHub Sponsors. Your support will help cover costs related to improving the codebase, adding new features, and providing better documentation. [![Sponsor](https://img.shields.io/badge/Sponsor-GitHub%20Sponsors-brightgreen)](https://github.com/sponsors/ipazc) -Your support will help cover costs related to improving the codebase, adding new features, and providing better documentation. + +## Acknowledgments + +This project has evolved over time with contributions from multiple developers. While the current codebase has been completely rewritten, we acknowledge and appreciate the valuable input and collaboration from past contributors. + +A special thanks to everyone who has submitted pull requests, reported issues, or provided feedback to make this project better. + +For a full list of contributors, please visit the [GitHub contributors page](https://github.com/ipazc/mtcnn/graphs/contributors). + ## License diff --git a/docs/references.md b/docs/references.md index 9cd40b6..a3df79d 100644 --- a/docs/references.md +++ b/docs/references.md @@ -1,58 +1,71 @@ # References -This document provides a detailed list of references, including the original research papers and projects that served as the foundation for this MTCNN implementation. Additionally, it includes information on how to properly cite this work if used in your research or projects. +This document provides references to the foundational work that inspired this implementation of the MTCNN (Multitask Cascaded Convolutional Networks) model, as well as guidelines on how to properly cite this library if used in your research or projects. -## Citation +## Citation for this MTCNN Library -If you use this library in your research or projects, please consider citing the original paper where the MTCNN model was introduced. This paper presents the Joint Face Detection and Alignment using Multitask Cascaded Convolutional Networks, a groundbreaking approach for face detection and landmark alignment. +If you use this implementation of the MTCNN library in your research or projects, please consider citing the following Zenodo entry: -### Original Paper: +```bibtex +@software{ivan_de_paz_centeno_2024_13901378, + author = {Iván de Paz Centeno}, + title = {ipazc/mtcnn: v1.0.0}, + month = oct, + year = 2024, + publisher = {Zenodo}, + version = {v1.0.0}, + doi = {10.5281/zenodo.13901378}, + url = {https://doi.org/10.5281/zenodo.13901378} +} +``` + +This citation provides proper credit for this specific implementation of the MTCNN library, as hosted on Zenodo. -- **Authors**: K. Zhang, Z. Zhang, Z. Li, and Y. Qiao +## Original Research Paper + +The development of the MTCNN model is based on the original paper titled **"Joint Face Detection and Alignment Using Multitask Cascaded Convolutional Networks"** by K. Zhang, Z. Zhang, Z. Li, and Y. Qiao. If your work benefits from the concepts or algorithms presented in this library, please consider citing this paper: + +- **Authors**: K. Zhang, Z. Zhang, Z. Li, Y. Qiao - **Journal**: IEEE Signal Processing Letters - **Title**: Joint Face Detection and Alignment Using Multitask Cascaded Convolutional Networks - **Year**: 2016 -- **Volume**: 23 -- **Number**: 10 -- **Pages**: 1499-1503 -- **Keywords**: Benchmark testing, Computer architecture, Convolution, Detectors, Face, Face detection, Training, Cascaded convolutional neural network (CNN), Face alignment - **DOI**: [10.1109/LSP.2016.2603342](https://doi.org/10.1109/LSP.2016.2603342) ```bibtex @article{7553523, - author={K. Zhang and Z. Zhang and Z. Li and Y. Qiao}, - journal={IEEE Signal Processing Letters}, - title={Joint Face Detection and Alignment Using Multitask Cascaded Convolutional Networks}, - year={2016}, - volume={23}, - number={10}, - pages={1499-1503}, - keywords={Benchmark testing;Computer architecture;Convolution;Detectors;Face;Face detection;Training;Cascaded convolutional neural network (CNN);face alignment;face detection}, - doi={10.1109/LSP.2016.2603342}, - ISSN={1070-9908}, - month={Oct} + author={K. Zhang and Z. Zhang and Z. Li and Y. Qiao}, + journal={IEEE Signal Processing Letters}, + title={Joint Face Detection and Alignment Using Multitask Cascaded Convolutional Networks}, + year={2016}, + volume={23}, + number={10}, + pages={1499-1503}, + doi={10.1109/LSP.2016.2603342}, + month={Oct} } ``` -### Abstract of the Original Paper: -The paper presents a multitask cascaded convolutional network (MTCNN) for joint face detection and alignment. This method integrates the detection and alignment process into a unified architecture, which significantly enhances the accuracy and speed of facial landmark localization. The system uses a three-stage network to predict face locations and landmarks iteratively, improving the results progressively across each stage. +### Summary of the Original Paper + +The paper introduces a multitask cascaded convolutional network (MTCNN) designed for joint face detection and alignment. This architecture improves accuracy and speed by integrating both tasks into a unified process, using a three-stage network to progressively refine facial landmark predictions. -## Original MTCNN Repository +## Original MTCNN Implementation -This library is based on the original implementation by Kaipeng Zhang, who made the pretrained networks and the code available for the research community. If you are using the models or weights provided in this library, you may also consider citing the original GitHub repository: +This library builds upon the original implementation by Kaipeng Zhang, who developed the MTCNN model and released the code for the research community. If you are using the models or weights provided in this library, you may also want to reference the original implementation: - **[Original MTCNN Implementation by Kaipeng Zhang](https://github.com/kpzhang93/MTCNN_face_detection_alignment/tree/master/code)** -This repository includes the source code, pre-trained weights, and additional information related to the original MTCNN framework, all released under the MIT license. +The original implementation includes the source code, pre-trained weights, and additional information related to the MTCNN framework. ## Related Work -This project also draws inspiration from the **FaceNet's MTCNN implementation** by David Sandberg. This implementation is part of a larger face recognition framework called FaceNet, which uses the MTCNN architecture to handle the task of face alignment prior to recognition. You may also want to refer to this project if you are using concepts from this work: +This project also draws inspiration from the **FaceNet's MTCNN implementation** by David Sandberg, which incorporates the MTCNN architecture into the FaceNet framework for face alignment prior to recognition. You may refer to this project if you use related concepts: - **[Facenet's MTCNN implementation](https://github.com/davidsandberg/facenet/tree/master/src/align)** +## Acknowledgments -## About this project +This implementation of the MTCNN library was developed by Iván de Paz Centeno, with contributions from various developers over the history of the project. Special thanks to all the contributors for their valuable input, which has helped improve the library. You can view the full list of contributors on the [GitHub contributors page](https://github.com/ipazc/mtcnn/graphs/contributors). -The code for this project was created to standardize face detection and provide an easy-to-use framework that helps the research community push the boundaries of AI knowledge. Learn more about the author of this code on [Iván de Paz Centeno's website](https://ipazc.com) +For more information about the author, visit [Iván de Paz Centeno's website](https://ipazc.com).