In the domain of facial recognition, our project stands out by integrating mask detection with deep learning-powered facial attribute analysis. We tackle the contemporary issue of obscured facial features due to mask-wearing, ensuring our algorithm remains effective in todayβs mask-prevalent society. Our solution, utilizing the advanced InceptionV3 architecture along with the comprehensive DeepFace library, is capable of identifying faces in a variety of settings, discerning masked from unmasked individuals, and further analyzing unmasked faces for age, gender, race, and emotion.
The key milestones in our project included:
- Mask Detection Model Development: Developing a model to identify if faces in images are masked, using our dataset and InceptionV3's transfer learning capabilities for enhanced accuracy.
- Facial Analysis with Mask Detection Integration: Combining mask detection with DeepFace-driven facial analysis to accurately annotate unmasked faces with attributes such as age, gender, race, and emotion.
- Real-World Application Demo: Showcasing the model's effectiveness in a real-world scenario, demonstrating facial analysis and mask detection on a sample video.
Facial Attribute Analysis with Mask Detection in Action:
Facial_Attribute_Analysis_with_Mask_Detection.ipynb
: The Jupyter notebook detailing our journey from data prep to model evaluation.test_image/
: Test image showcasing the model's generalization capability.README.md
: You're currently reading this file! Provides an overview and useful information about the project.
- Clone this Repository: First, clone this repo to your local machine.
- Download the Dataset: Access the dataset for mask detection from Kaggle. Download and extract it to a known directory on your machine.
- Update Dataset Path: Open the
Facial_Attribute_Analysis_with_Mask_Detection.ipynb
notebook and update the dataset path to the location where you extracted the dataset. - Setup Dependencies: Ensure that all necessary Python libraries are installed.
- Execution: Run all cells in the notebook to see the results.
- π Kaggle Notebook: Interested in a Kaggle environment? Explore the notebook here.
- πΉ Input Video Data: Access the raw and modified video here.
- π₯ Project Demo: Watch the live demonstration of this project on YouTube.
- π€ Connect on LinkedIn: Have questions or looking for collaboration? Let's connect on LinkedIn.