Datasets:
license: cc-by-nc-nd-4.0
tags:
- video
- image
- web-camera
- Facial Recognition
- Computer Vision
- Machine learning
- Security
task_categories:
- image-segmentation
size_categories:
- 10K<n<100K
Web Camera People Behavior Dataset for computer vision tasks
Dataset includes 2,300+ individuals, contributing to a total of 53,800+ videos and 9,300+ images captured via webcams. It is designed to study social interactions and behaviors in various remote meetings, including video calls, video conferencing, and online meetings.
By leveraging this dataset, developers and researchers can enhance their understanding of human behavior in digital communication settings, contributing to advancements in technology and software designed for remote collaboration. - Get the data
Metadata for the dataset
.png?generation=1743156643952762&alt=media)
Dataset boasts an impressive >97% accuracy in action recognition (including actions such as sitting, typing, and gesturing) and ≥97% precision in action labeling, making it a highly reliable resource for studying human behavior in webcam settings.
Frequently Asked Questions
How was the data collected?
The recordings were collected through a standardized acquisition process using both webcams and mobile phones. Capturing the same participant with different devices provides consistent cross-device facial data for evaluating the robustness of biometric and computer vision models.
What metadata is included with the dataset?
Each recording includes metadata such as participant ID, gender, ethnicity, smartphone type, and webcam model. These annotations enable researchers to analyze model performance across different demographics and hardware configurations while supporting fairness and robustness evaluations.
💵 Buy the Dataset: This is a limited preview of the data. To access the full dataset, please contact us at https://unidata.pro to discuss your requirements and pricing options.
Researchers can utilize this dataset to explore the impacts of web cameras on social and professional interactions, as well as to study the security features and audio quality associated with video streams. The dataset is particularly valuable for examining the nuances of remote working and the challenges faced during video conferences, including issues related to video quality and camera usage.