---
title: "Anti-Spoofing Real Videos Dataset"
description: "98,000+ files 50,000+ people 1 selfie and 1 video of each person"
url: "https://unidata.pro/datasets/face-anti-spoofing/"
date_modified: "2026-03-23T13:30:25+03:00"
language: "en-US"
---
The dataset comprises anti-spoofing real videos with paired images of human faces, featuring diverse resolutions, demographics, and metadata to support face anti-spoofing research, spoofing detection, liveness verification, and the development of biometric authentication systems against presentation attacks

## Dataset Structure

### The Numbers Section

**Numbered list:**

- **Number:** 87,340 — **Text:** Files
- **Number:** 179 — **Text:** Countries
- **Number:** 43,670 — **Text:** People

### Tooltips Section

**Tooltip items:**

- **Name:** Facial Recognition
- **Name:** iBeta
- **Name:** Liveness Detection
- **Name:** Security
- **Name:** Anti-spoofing

### Dataset Information

**Table with data:**

| Characteristic | Data |
| --- | --- |
| Description | Live videos of people for anti-spoofing tasks |
| Data types | Image, Video |
| Tasks | Face recognition, face detection |
| Number of people | 43,670 |
| Number of files in a set | 2 (1 image & 1 video) |
| Total number of video | 87,340 |
| Labeling | Only technical characteristics and metadata (age, gender, ethnicity) |
| Gender | Male, Female |
| Ethnicity | Asian (30%), African (70%) |
| Age | Min = 18, max = 80, mean = 45 |
| Number of countries | 179 |

**Media Slider:**

- **Video on Slayder:** <https://unidata.pro/wp-content/uploads/2024/06/anti-spoofing-real.mp4>
- **Image in the slider:** ![](https://unidata.pro/wp-content/uploads/2024/06/selfie.webp)

**Link to the sample:** [Download sample](https://drive.google.com/drive/folders/1jehloN3IZ8uvnX-rPVMG1rAu4UJjRxg8)

### LLM Languages

**Section Title:** Statistics

**List of Statistics:**

- **Filter by:** Age Distribution — **GIF image:** Age Distribution — **Table with data:**

| Age | Count |
| --- | --- |
| Under 18 | 2106 |
| 19 - 25 | 20955 |
| 26 - 32 | 12216 |
| 33 - 39 | 5756 |
| 40 - 46 | 2356 |
| 47 - 53 | 872 |
| 54 - 59 | 256 |
| 60 - 66 | 109 |
| 67+ | 44 |
- **Filter by:** Phone Brand — **GIF image:** Phone Brand — **Table with data:**

| Brand | Count |
| --- | --- |
| Apple | 5,305 |
| Samsung | 4,884 |
| Xiaomi | 2,548 |
| Huawei | 1,345 |
| Infinix | 1,339 |
| Android | 1,159 |
| Tecno | 1,067 |
| Vivo | 1,021 |
| Oppo | 1,006 |
| Realme | 735 |
| Motorola | 678 |
| Nokia | 279 |
| Itel | 236 |
| LG | 188 |
| ZTE | 108 |
| Alcatel | 86 |

### Statistics - Charts

**Charts with Titles:**

- **Shortcode:** [ays_chart id='49'] — **caption above the graph:** Gender distribution
- **Shortcode:** [ays_chart id='51'] — **caption above the graph:** Ethnicity distribution
- **Shortcode:** [ays_chart id='50'] — **caption above the graph:** Resolution distribution
- **Shortcode:** [ays_chart id='52'] — **caption above the graph:** Continent distribution

### Technical Specifications

**Table with data:**

| Characteristic | Data |
| --- | --- |
| Video Extensions | mp4, MOV |
| Image Extension | jpg |
| Video Resolutions | 1920 x 1080p, 480 x 360p, 1280 x 720p, 720 x 480p, 640 x 480p, 1920 x 920p |
| Video Duration | Mean = 9, median = 9, min = 2, max = 34 |
| Frames per second | Mean = 26.6 |
| Devices | iPhone 13 (30%), Google Pixel (70%) |

**Source and data collection methodology:** Data source. Data was collected via crowdsourcing platforms.

### Dataset Use Cases - Slider

**Industry Cards:**

- **Industry:** Financial Services & Digital Banking — **Title:** Preventing Facial Fraud in Biometric Authentication — **Text:** Banks and fintech platforms rely on facial recognition for secure onboarding and identity verification. This face anti-spoofing dataset provides real video footage of spoofing attacks, such as photo attacks, 3D masks, and printed photos, helping improve liveness detection and prevent biometric fraud across online authentication systems.
- **Industry:** Mobile Devices & Consumer Electronics — **Title:** Enhancing Smartphone Unlock Security — **Text:** Face unlock features are often targets for spoofing attempts. With real-world samples of anti-spoofing techniques, this dataset allows manufacturers to train anti-spoofing algorithms that distinguish between real faces and fake faces. It supports better liveness detection and anti-spoofing systems for everyday biometric security on consumer devices.
- **Industry:** Government & Border Control — **Title:** Securing Access with Strong Liveness Detection — **Text:** This dataset helps improve biometric authentication systems used in airports, border control, and secure facilities. By providing real examples of presentation attacks and spoofing techniques, it supports the development of accurate anti-spoofing technology that can reliably detect real persons during identity verification at checkpoints and kiosks.
- **Industry:** Research & Development — **Title:** Training Robust Anti-Spoofing Models — **Text:** This dataset is ideal for training deep learning models in academic and industry research. With high-quality video samples of spoofing detection scenarios, it helps test and refine state-of-the-art methods, including convolutional networks, texture analysis, and facial movements – pushing forward the development of more accurate anti-spoofing solutions.

### Fact

**FAQs Heading:** FAQs

**List of Questions:**

- **Question:** What is Anti-Spoofing Real Videos Dataset used for? — **Answer:** The dataset is designed to develop and evaluate anti-spoofing algorithms, detection techniques, and biometric authentication systems. It supports face recognition, liveness detection, and the prevention of spoofing attacks such as photo attacks, 3D masks, and print attacks in modern biometric security systems.
- **Question:** What types of annotations are provided? — **Answer:** The dataset is labeled with technical characteristics and metadata, including participant age, gender, and ethnicity. These annotations help create structured training datasets for spoofing detection and facial recognition research.
- **Question:** Is this dataset real or synthetic? — **Answer:** The face anti-spoofing dataset consists of real videos and images of genuine participants, not synthetic or artificially generated. This ensures reliable training data for testing anti-spoofing technology against known attacks and enhancing biometric security.
- **Question:** How long does it take to receive the dataset? — **Answer:** Once you submit a request, we will reach out to review the details and complete the necessary documents. After signing and payment, the dataset will be delivered within 3-10 days, allowing you to begin biometrics testing, face recognition, liveness detection, and attack detection in biometric systems promptly.
- **Question:** How are Unidata datasets licensed? — **Answer:** Unidata datasets follow a dual licensing model: evaluation samples are provided for free, while the full dataset is accessible only by purchase.
- **Question:** Do Unidata datasets follow GDPR or other data privacy regulations? — **Answer:** Yes. Each dataset follows GDPR guidelines and complies with applicable data protection laws. Data is carefully collected from legal sources to ensure responsible usage.
- **Question:** How are Unidata datasets stored? — **Answer:** Unidata secures all datasets within AWS’s cloud platform, built for flexibility and continuous availability. We adhere to ISO 27001 and ISO 27701 certifications, which uphold the highest standards in information security and privacy management. This commitment provides our clients with confidence in data safety and reliability.
- **Question:** What technical characteristics are included in the dataset? — **Answer:** The dataset provides JPG images together with MP4 and MOV videos captured at multiple resolutions, including 1920 × 1080, 1280 × 720, and 640 × 480. Videos range from 2 to 34 seconds in length, with an average duration of 9 seconds and a mean frame rate of 26.6 FPS, making them suitable for training real-world face anti-spoofing models.
- **Question:** Why is worldwide participant diversity important for anti-spoofing? — **Answer:** With contributors from 179 countries, the dataset reflects a broad range of facial appearances and recording conditions encountered in global biometric applications. This diversity helps improve the robustness of face anti-spoofing and identity verification systems deployed internationally.

[Full list of this site's AI-readable pages](https://unidata.pro/llms.txt)