--- title: "Latex Mask Attacks Dataset" description: "This is a high-resolution collection of 11,100 videos of people wearing latex masks, created for training facial recognition and spoofing detection models. Featuring diverse facial…" url: "https://unidata.pro/datasets/latex-mask-attacks/" date_modified: "2025-10-08T16:54:25+03:00" language: "en-US" --- This is a high-resolution collection of 11,100 videos of people wearing latex masks, created for training facial recognition and spoofing detection models. Featuring diverse facial features, metadata, and realistic presentation attacks, it supports fraud prevention research and iBeta Level 2 certification for robust biometric security systems. ## Dataset Structure ### The Numbers Section **Numbered list:** - **Number:** 11,100+ — **Text:** videos - **Number:** 5 — **Text:** devices ### Tooltips Section **Tooltip items:** - **Name:** iBeta - **Name:** Liveness Detection - **Name:** Computer Vision - **Name:** Security - **Name:** Facial Recognition - **Name:** Anti-spoofing ### Dataset Information **Table with data:** | Characteristic | Data | | --- | --- | | Description | Video of people in latex masks training algorithms to detect biometric hacking attempts. | | Data types | Video | | Tasks | Face recognition, Computer Vision | | Total number of files | 11,100 | | Labeling | Only technical characteristics and metadata (age, gender, ethnicity) | | Gender | Male, Female | | Number of attributes | 31 | **Media Slider:** - **Video on Slayder:** - **Video on Slayder:** **Link to the sample:** [Download sample](https://drive.google.com/drive/folders/1mYLls4KY12jCpzlnH3Jl0fv6gqJn5GM2) ### Statistics - Charts **Charts with Titles:** - **Shortcode:** [ays_chart id="37"] — **caption above the graph:** Devices in the dataset - **Shortcode:** [ays_chart id="38"] — **caption above the graph:** Gender distribution ### Technical Specifications **Table with data:** | Characteristic | Data | | --- | --- | | Video extension | mp4 | | Video Resolutions | Min = 1920х1080, Max = 3840х2160 | | Video duration | 1-2 second | | Number of background | 9 | | Video overlap | No more than 5% | | Devices | IOS, Android | **Source and data collection methodology:** Source and collection methodology. Data was collected via crowdsourcing platforms. ### Dataset Use Cases - Slider **Industry Cards:** - **Industry:** Biometric Security — **Title:** Strengthening Spoofing Detection Systems — **Text:** Latex Mask Attacks Dataset provides 11,100 high-resolution videos of human faces with latex masks, helping researchers build stronger spoofing detection and attack detection systems. This dataset supports the training of recognition algorithms that distinguish between real faces and 3D masks, reducing risks from advanced presentation attacks. - **Industry:** Financial Services — **Title:** Enhancing Fraud Prevention in Authentication — **Text:** Banks and fintech platforms use such datasets to improve facial recognition during identity verification and KYC checks. With datasets containing face images and spoofing attacks, institutions can train deep models to detect mask attacks and ensure secure, fraud-resistant systems for financial transactions and digital onboarding. - **Industry:** Healthcare & Access Control — **Title:** Securing Medical Data and Restricted Areas — **Text:** Hospitals and laboratories benefit from this dataset by developing training data for recognition systems that protect medical images and records. With videos of latex masks simulating face presentation attacks, developers can build detection algorithms that prevent unauthorized entry and safeguard sensitive health information against biometric spoofing. - **Industry:** AI Research & Certification — **Title:** Advancing Deep Learning and iBeta Standards — **Text:** For AI researchers, Latex Mask Attacks Dataset, comprising videos of latex mask attacks, offers valuable material for deep learning models in facial recognition. By analyzing facial features of mask attacks, teams can create learning algorithms that improve image classification and prepare systems for compliance with iBeta Level 2 certification and other security benchmarks. ### Fact **FAQs Heading:** FAQs **List of Questions:** - **Question:** What types of annotations are provided? — **Answer:** The dataset provides metadata annotations, including demographic details and technical recording attributes. These labels help train deep learning models for identifying real faces vs. latex masks. - **Question:** How was the dataset collected? — **Answer:** Videos were captured with binocular RGB and infrared cameras on iOS and Android devices. Recordings span nine different backgrounds with resolutions ranging from 1920×1080 to 3840×2160, ensuring robust training data for detection algorithms. - **Question:** What is the average duration and quality of videos? — **Answer:** Each video is 1–2 seconds long, designed for quick facial recognition testing. The dataset includes high-resolution footage suitable for object detection, image classification, and biometric attack research. - **Question:** What types of attacks are included in Latex Mask Attacks Dataset? — **Answer:** The dataset focuses on latex mask presentation attacks, where individuals wear realistic facial masks to imitate another person’s identity. These attack scenarios are valuable for evaluating liveness detection, presentation attack detection (PAD), and face recognition algorithms under real-world biometric threats. - **Question:** How does this dataset differ from other spoofing datasets? — **Answer:** Unlike datasets focusing on printed photos or replay attacks, this collection highlights latex masks and 3D presentation attacks. It provides more advanced spoofing scenarios, making it essential for next-generation liveness detection systems. - **Question:** How are Unidata datasets licensed? — **Answer:** Unidata datasets follow a dual-licensing model. Free dataset samples are provided for research trials and algorithm testing, while full, high-resolution datasets - including Latex Mask Attacks Dataset - are available exclusively through purchase. - **Question:** Do Unidata datasets follow GDPR or other data privacy regulations? — **Answer:** Yes. All Unidata datasets are curated in full compliance with GDPR and international data protection regulations. - **Question:** How long does it take to receive the dataset? — **Answer:** Once your request is submitted, our team will review the details, prepare the necessary documentation, and provide an agreement for signature. After payment confirmation, the dataset will be delivered within 3 to 10 business days via secure digital transfer. - **Question:** Is this a real-world dataset or synthetic data? — **Answer:** Latex Mask Attacks Dataset consists of real-world video recordings collected through crowdsourcing platforms, not synthetic data. Each clip captures realistic mask attack scenarios under different environments and devices, making it valuable for developing robust facial recognition and biometric attack detection algorithms. [Full list of this site's AI-readable pages](https://unidata.pro/llms.txt)