In this blog, we will follow a structured approach to build and use a Face Recognition Dataset. The workflow includes:<\/p>\n\n\n\n
- \n
- Exploring Popular Public Datasets<\/strong> \u2013 Understanding datasets like LFW, VGGFace2, and CelebA, which help benchmark models and give context.<\/li>\n\n\n\n
- Creating Your Own Custom Dataset<\/strong> \u2013 Capturing images of individuals using a webcam and organizing them into a structured folder system.<\/li>\n\n\n\n
- Setting Up the Environment<\/strong> \u2013 Installing and configuring Python libraries such as OpenCV, face_recognition, NumPy, and Matplotlib.<\/li>\n\n\n\n
- Loading and Exploring the Dataset<\/strong> \u2013 Loading images from both custom and public datasets, visualizing sample images, checking labels, dimensions, and overall dataset structure to ensure everything is ready for preprocessing.<\/li>\n\n\n\n
- Preprocessing the Dataset<\/strong> \u2013 Converting images to grayscale, resizing, and normalizing them for model training.<\/li>\n\n\n\n
- Training a Face Recognition Model<\/strong> \u2013 Encoding faces into numeric vectors and teaching the model to recognize them.<\/li>\n\n\n\n
- Testing and Evaluating the Model<\/strong> \u2013 Using both custom and public datasets to check accuracy and performance.<\/li>\n\n\n\n
- Augmenting the Dataset<\/strong> \u2013 Improving model performance with image transformations such as flips, rotations, and zooms.<\/li>\n<\/ol>\n\n\n\n
This step-by-step workflow ensures that even beginners can understand, create, and implement a Face Recognition Dataset efficiently.<\/p>\n\n\n\n
Do check out HCL GUVI\u2019s Data Science eBook<\/a>, which gives you a clear, structured overview of how to collect, clean, and prepare image data for machine learning. It also walks you through essential concepts like data preprocessing, model training, and evaluation \u2014 all explained through beginner-friendly examples.<\/p>\n\n\n\n
1. Popular Face Recognition Datasets<\/strong><\/h2>\n\n\n\n
![\"](\"https:\/\/www.guvi.in\/blog\/wp-content\/uploads\/2025\/12\/1-1-1.png\" "\\"\\"")
<\/figure>\n\n\n\nBefore creating your own Face Recognition Dataset<\/strong>, it is helpful to know about the widely used public datasets in face recognition. These datasets provide a large collection of images that can be used for training, testing, and benchmarking models. Here are some of the most popular ones:<\/p>\n\n\n\n
- \n
- LFW (Labeled Faces in the Wild):<\/strong> This dataset contains over 13,000 images of faces collected from the internet. Each image is labeled with the person\u2019s name, making it ideal for testing face verification and recognition algorithms. LFW is especially useful for beginners who want a simple, real-world dataset to practice on. Download the dataset here – Kaggle
<\/a><\/li>\n\n\n\n- VGGFace2:<\/strong> VGGFace2 has 3.3 million images of more than 9,000 people. It includes faces under different poses, lighting conditions, and ages. This variety makes it perfect for building robust face recognition models that can handle real-world variations. Download the dataset here – Kaggle
<\/a><\/li>\n\n\n\n- CelebA:<\/strong> With around 200,000 images of celebrities, CelebA not only provides face images but also includes 40 facial attributes such as glasses, smiling, or gender. This dataset is useful if you want to train models for both face recognition and facial attribute detection. Download the dataset here – Kaggle
<\/a><\/li>\n\n\n\n- CASIA-WebFace:<\/strong> CASIA-WebFace consists of 494,414 images of 10,575 individuals. It is widely used in research for large-scale face recognition projects and provides a good balance between dataset size and diversity. Download the dataset here – Kaggle<\/a>
<\/li>\n\n\n\n- MS-Celeb-1M:<\/strong> This massive dataset contains 10 million images of 100,000 identities. It is designed for large-scale face recognition and can help train high-performance models, but it requires significant computational resources to handle. Download the dataset here – MS-Celeb<\/a><\/li>\n<\/ul>\n\n\n\n
While these public datasets are excellent for learning and benchmarking, creating your own custom Face Recognition Dataset becomes important when you need a system tailored to your specific scenario, like an office attendance system, classroom monitoring, or personalized authentication. A custom dataset ensures the model learns the exact faces it will encounter, which improves accuracy in practical applications.<\/p>\n\n\n\n
2. Creating Your Own Face Recognition Dataset<\/strong><\/h2>\n\n\n\n
![\"Infographic](\"https:\/\/www.guvi.in\/blog\/wp-content\/uploads\/2025\/12\/2-2-1.png\" "\\"\\"")
<\/figure>\n\n\n\nCreating a custom Face Recognition Dataset allows you to train a model that works for your specific environment. Custom datasets are important when you need a system to recognize specific people, such as for office attendance, classroom monitoring, or personal authentication.<\/p>\n\n\n\n
Here\u2019s a simple way to capture faces using Python<\/a> and OpenCV.<\/p>\n\n\n\n
Step 2.1: Install Required Libraries<\/strong><\/h3>\n\n\n\n
Purpose:<\/strong> These libraries allow you to access your webcam, process images, and store them in an organized way.<\/p>\n\n\n\n
pip install opencv-python\npip install numpy\n<\/code><\/pre>\n\n\n\nOpenCV:<\/strong> Captures images from your webcam and processes them.<\/p>\n\n\n\n
NumPy:<\/strong> Handles image data in arrays for easy manipulation.<\/p>\n\n\n\n
Step 2.2: Set Up the Dataset Folder<\/strong><\/h3>\n\n\n\n
Purpose:<\/strong> Organize images for each person separately so your model can easily identify them later.<\/p>\n\n\n\n
import os\n\ndataset_path = \"dataset\"\nif not os.path.exists(dataset_path):\n os.makedirs(dataset_path)\n\nperson_name = input(\"Enter the name of the person: \")\nperson_path = os.path.join(dataset_path, person_name)\nif not os.path.exists(person_path):\n os.makedirs(person_path)\n<\/code><\/pre>\n\n\n\nExplanation:<\/strong><\/p>\n\n\n\n
- \n
- Creates a main folder named as dataset<\/li>\n\n\n\n
- Each person has a subfolder named after them<\/li>\n\n\n\n
- Keeps your Face Recognition Dataset structured and easy to manage<\/li>\n<\/ul>\n\n\n\n
Step 2.3: Capture Faces from Webcam<\/strong><\/h3>\n\n\n\n
Purpose:<\/strong> Collect multiple images of each person under different conditions to improve model accuracy.<\/p>\n\n\n\n
import cv2\n# Initialize webcam\ncap = cv2.VideoCapture(0)\nface_cascade = cv2.CascadeClassifier(cv2.data.haarcascades + \"haarcascade_frontalface_default.xml\")\ncount = 0\nwhile True:\n ret, frame = cap.read()\n if not ret:\n break\n gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)\n faces = face_cascade.detectMultiScale(gray, 1.3, 5)\n for (x, y, w, h) in faces:\n count += 1\n face_img = gray[y:y+h, x:x+w]\n cv2.imwrite(f\"{person_path}\/{person_name}_{count}.jpg\", face_img)\n cv2.rectangle(frame, (x, y), (x+w, y+h), (255, 0, 0), 2)\n cv2.imshow(\"Capturing Faces\", frame)\n if cv2.waitKey(1) & 0xFF == ord('q') or count >= 50:\n break\ncap.release()\ncv2.destroyAllWindows()\nprint(f\"Collected {count} images for {person_name}\")\n<\/code><\/pre>\n\n\n\nExplanation:<\/strong><\/p>\n\n\n\n
- \n
- Uses OpenCV to capture live video from the webcam<\/li>\n\n\n\n
- Detects faces using a Haar Cascade classifier<\/li>\n\n\n\n
- Saves 50 face images per person by default<\/li>\n\n\n\n
- Draws a rectangle around the detected face for visual feedback<\/li>\n<\/ul>\n\n\n\n
Tip:<\/strong> Capture faces with different angles, expressions, and lighting conditions to create a robust Face Recognition Dataset.<\/p>\n\n\n\n
Folder Structure:<\/strong><\/p>\n\n\n\n
dataset\/\n \u251c\u2500\u2500 person1\/\n \u2502 \u251c\u2500\u2500 person1_1.jpg\n \u2502 \u251c\u2500\u2500 person1_2.jpg\n \u2502 \u2514\u2500\u2500 ...\n \u2514\u2500\u2500 person2\/\n \u251c\u2500\u2500 person2_1.jpg\n \u251c\u2500\u2500 person2_2.jpg\n \u2514\u2500\u2500 ...\n<\/code><\/pre>\n\n\n\n3. Setting Up Your Environment<\/strong><\/h2>\n\n\n\n
![\"Diagram](\"https:\/\/www.guvi.in\/blog\/wp-content\/uploads\/2025\/12\/3-2-1.png\" "\\"\\"")
<\/figure>\n\n\n\nBefore using your custom Face Recognition Dataset or any public datasets, you need to set up a Python environment with the necessary libraries<\/a>. This ensures that your model can capture, process, and recognize faces efficiently.<\/p>\n\n\n\n
Step 3.1: Install Required Libraries<\/strong><\/h3>\n\n\n\n
pip install face_recognition\npip install matplotlib\npip install scikit-learn\n<\/code><\/pre>\n\n\n\nExplanation:<\/strong><\/p>\n\n\n\n
- VGGFace2:<\/strong> VGGFace2 has 3.3 million images of more than 9,000 people. It includes faces under different poses, lighting conditions, and ages. This variety makes it perfect for building robust face recognition models that can handle real-world variations. Download the dataset here – Kaggle
- Creating Your Own Custom Dataset<\/strong> \u2013 Capturing images of individuals using a webcam and organizing them into a structured folder system.<\/li>\n\n\n\n
![\"Infographic](\"https:\/\/www.guvi.in\/blog\/wp-content\/uploads\/2025\/12\/4-8.png\" "\\"\\"")
<\/figure>\n\n\n\n![\"Infographic](\"https:\/\/www.guvi.in\/blog\/wp-content\/uploads\/2025\/12\/5-2-1.png\" "\\"\\"")
<\/figure>\n\n\n\n![\"Diagram](\"https:\/\/www.guvi.in\/blog\/wp-content\/uploads\/2025\/12\/6-2-1.png\" "\\"\\"")
<\/figure>\n\n\n\n![\"](\"https:\/\/www.guvi.in\/blog\/wp-content\/uploads\/2025\/12\/7-1-1.png\" "\\"\\"")
<\/figure>\n\n\n\n