Face Data (Detection)

## Overview

Face Data (Detection) is a dataset for object detection tasks - it contains 001 002 003 004 005 006 007 annotations for 6,672 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

Dataset

This dataset was created by dAReDeViL555

Contents

The 399 Asians - 35,112 Images Multi-pose Face Data with 21 Facial Landmarks Annotation data is collected from 399 people. The data diversity includes multiple poses, different ages, different light conditions and multiple scenes. This data can be used for tasks such as face detection and face recognition. Thee accuracy of labels of gender, face pose, year of birth, light condition, scene and wearing glasses or not is more than 97%;annotation accuracy of facial landmarks is more than 97%

Dataset Card for "Control-Face-data"

More Information needed

Biometric Data

FileMarket provides a comprehensive Biometric Data set, ideal for enhancing AI applications in security, identity verification, and more. In addition to Biometric Data, we offer specialized datasets across Object Detection Data, Machine Learning (ML) Data, Large Language Model (LLM) Data, and Deep Learning (DL) Data. Each dataset is meticulously crafted to support the development of cutting-edge AI models.

Data Size: 20,000 IDs

Race Distribution: The dataset encompasses individuals from diverse racial backgrounds, including Black, Caucasian, Indian, and Asian groups.

Gender Distribution: The dataset equally represents all genders, ensuring a balanced and inclusive collection.

Age Distribution: The data spans a broad age range, including young, middle-aged, and senior individuals, providing comprehensive age coverage.

Collection Environment: Data has been gathered in both indoor and outdoor environments, ensuring variety and relevance for real-world applications.

Data Diversity: This dataset includes a rich variety of face poses, racial backgrounds, age groups, lighting conditions, and scenes, making it ideal for robust biometric model training.

Device: All data has been collected using mobile phones, reflecting common real-world usage scenarios.

Data Format: The data is provided in .jpg and .png formats, ensuring compatibility with various processing tools and systems.

Accuracy: The labels for face pose, race, gender, and age are highly accurate, exceeding 95%, making this dataset reliable for training high-performance biometric models.

Face Mask Detection Dataset

Overview

This dataset is designed for developing and training face mask detection models that can generalize well across real-world variations. Unlike typical mask datasets that only distinguish between “mask” and “no mask,” this collection includes five diverse classes to handle complex, real-world scenarios where faces may be partially covered or occluded.

Purpose **Real-world face detection systems often misclassify blocked or partially visible faces as “masked.” This dataset addresses that by introducing additional contextual classes such as beard and face_blocked to help models learn nuanced visual differences between intentional coverings (masks) and unintentional occlusions.

Use Cases

Training CNNs, MobileNet, EfficientNet, or YOLO models for face mask detection. Fine-tuning pre-trained models for COVID-19 compliance monitoring or smart surveillance systems. Benchmarking robustness of face detection algorithms in occluded or cluttered scenes.

Data Details

Each image is annotated with one of the five labels. The dataset includes variations in lighting, angles, facial hair, and occlusion. Suitable for both classification and object detection tasks.

1,507 People 102,476 Images Multi-pose and Multi-expression Face Data. The data includes 1,507 Asians (762 males, 745 females). For each subject, 62 multi-pose face images and 6 multi-expression face images were collected. The data diversity includes multiple angles, multiple poses and multple light conditions image data from all ages. This data can be used for tasks such as face recognition and facial expression recognition.

Off-the-shelf biometric data (human face) covers 3D depth, segmentation: face organs and accessory, key points, facial expression, alpha Matte, age in variety and etc. All the Biometric Data are collected with signed authorization agreement.

Mahinur/face-data dataset hosted on Hugging Face and contributed by the HF Datasets community

1. Specifications Product : Biometric Data

Data size : 200,000 ID

Race distribution : black people, Caucasian people, brown(Mexican) people, Indian people and Asian people

Gender distribution : gender balance

Age distribution : young, midlife and senior

Collecting environment : including indoor and outdoor scenes

Data diversity : different face poses, races, ages, light conditions and scenes Device : cellphone

Data format : .jpg/png

Accuracy : the accuracy of labels of face pose, race, gender and age are more than 97%

1. About Nexdata Nexdata owns off-the-shelf PB-level Large Language Model(LLM) Data, 3 million hours of Audio Data and 800TB of Annotated Imagery Data. These ready-to-go Biometric Data support instant delivery, quickly improve the accuracy of AI models. For more details, please visit us at https://www.nexdata.ai/datasets/computervision?source=Datarade

Dataset

This dataset was created by Bhargavi

Contents

50 people – 3D scanning face data. The collection scene is indoor. The data covers males and females. The age distribution ranges from youth to old age, mainly young and middle-aged. The collection device is a special scanner. The data can be used for tasks such as 3D face recognition, 3D face modeling, etc.

eliasfiz/idling-video-face-data-tokenised dataset hosted on Hugging Face and contributed by the HF Datasets community

Description

Contains face videos (.avi files) from 200 data subjects, acquired using Monochrome Camera MSC2-M42-1-A from Spectral Devices Inc.

During the recording, the data subjects were seated in the driver's seat of a vehicle, with the camera positioned on the dashboard behind the steering wheel.

The videos were acquired in the following scenarios, the aim of which was to incorporate different variabilities into the recorded face data:

(1) Indoors: The car was parked inside a garage, with controlled (artificial) lighting.

• Subject wore no deliberate accessories.
• Subject wore a hygienic mask.
• Subject wore a hat (cap).

(2) Outdoors: The car was parked outside, with uncontrolled (natural) lighting.

• Subject wore no deliberate accessories.
• Subject wore sunglasses.
• Subject wore a hat (cap).

For each of the acquisition scenarios described above, face videos of the data subject were captured while the subject:

(a) Remained still, with a neutral facial expression (5 seconds).
(b) Remained still, with a neutral facial expression, and with eyes closed (5 seconds).
(c) Performed natural face/head movements, e.g., turning, talking, smiling, laughing (15 seconds).

So in total, this dataset consists of 18 face videos per data subject. This amounts to 3,600 face videos for all 200 data subjects.

Reference

If you use this dataset, please cite the following publication:

V. Krivokuca Hahn, J. Maceiras, A. Komaty, P. Abbet and S. Marcel, 2024. "in-Car Biometrics (iCarB) Datasets for Driver Recognition: Face, Fingerprint, and Voice". arXiv:2411.17305, doi: https://doi.org/10.48550/arXiv.2411.17305.

I was working on the UTK Dataset and I been through many datasets full of images in different folders which requires a lot of data cleaning and preprocessing. So I tried to create this dataset in a more simplified manner keeping all the data in the form of a CSV and making it available to everyone.

Content

This dataset includes a CSV of facial images that are labeled on the basis of age, gender, and ethnicity. The dataset includes 27305 rows and 5 columns.

Acknowledgements

I downloaded the initial jpeg files from Kaggle

Inspiration

I hope many people use this dataset to create good CNN architectures in future (also it would help me learn more of deep learning too) PEACE!

Celeb Faces

## Overview

Celeb Faces is a dataset for object detection tasks - it contains Faces annotations for 423 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

FFHQ Faces Data Set

Ported to Kaggle from: https://github.com/NVlabs/ffhq-dataset

Related paper:
A Style-Based Generator Architecture for Generative Adversarial Networks
Tero Karras (NVIDIA), Samuli Laine (NVIDIA), Timo Aila (NVIDIA)
http://stylegan.xyz/paper

This data set only includes 'thumbnail' images (128px by 128px) due to size limitations

This table contains the data extraction results for the study Machine-based Stereotypes: How Machine Learning Algorithms Evaluate Ethnicity from Face Data. It contains 24 columns and 74 rows.

Abstract: Previous studies have found that it is more difficult to identify an emotional facial expression displayed by an older than a younger face. It is unknown whether this is caused by age-related changes such as wrinkles and folds interfering with perception, or by the aging of facial muscle, potentially reducing the ability of older individuals to display an interpretable expression. To discriminate between these two possibilities, we conducted a psychophysics experiment where participants attempted to identify emotional facial expression under different conditions. To control for the variables (wrinkles/folds vs facial muscles, we made use of Generative Adversarial Networks (GAN) to make images of faces look older or younger. As expected, emotions expressed by older faces (Condition 2) were harder to identify than those expressed by younger faces (Condition 1). Interestingly, participants' accuracy in identifying emotions was not affected when the "young faces" (Condition 1) were artificially aged (Condition 3). On the other hand, using a reverse aging filter to make the older faces (Condition 2) look young (Condition 4) significantly reduced the ability of our participants to identify the correct emotional expression.

Taken together, these results suggest that an age-related decline in ability to produce recognizable facial expressions, rather than the age-related physical changes in the face such as folds and wrinkles, explain why it is more difficult to recognize facial expressions from older faces. Consequently, facial muscle exercises might improve the capacity to convey facial emotional expressions in the elderly.To promote transparency and repeatability of our manuscript, "Having difficulties reading the facial expression of older individuals? Blame it on the facial muscles, not the wrinkles." currently under review for publications in Frontiers Psychology, we make the following files available: "facedata.xlsx", which contains the raw data collected for our study (400 trials x 28 participants = 11,200 trials) and an expanded table based upon the hierarchical logistic regression analysis (a more limited version will be available in the published manuscript).For questions about the raw data or table, please contact Nicolas Brunet at brunenm@millsaps.edu

Dataset

This dataset was created by Yash Chordia0

Released under MIT

Contents