gloriacao/Datachallenge_2

Dataset Card for Elderly Fallen People Data Set

The dataset includes falls and non falls images in indoor environments.

Dataset Details

Dataset Description

The dataset contains a total of 413 images of falls and non falls of elderly people in real-life scenarios (sitting, laying in bed or on the sofa, standing up, etc.). The images capture different angles, fall types, states of dress, lighting, etc. Participants are all white. The falls are simulated by 3 volunteers over 65 years old, 2 women and 1 man. The set contains 182 non-fall and 192 fall images in the training set, whereas 10 non-fall and 26 fall images in the validation set.

• Curated by: S. Maldonado-Bascón, C. Iglesias-Iglesias, P. Martín-Martín, S. Lafuente-Arroyo.

• License: Creative Commons Attribution 4.0 (CC-BY-4.0)

Uses

Direct Use

The dataset is used for training an image classifier model to distinguish falls and non falls, detecting where intervention is necessary or not. It is divided into training, validation and test set.

Out-of-Scope Use

Test set images must not be used for training systems, whereas training and validation set can be used for that scope.

Dataset Structure

The dataset is divided into three different datasets: training set, validation set and test set. The elderly set specifically contains 413 annotated images of 3 different volunteers over 65 years old.

Personal and Sensitive Information

The dataset doesn't include personally identifiable information, but it contais images of individuals in their home, which may be considered as sensitive visual data.

Bias, Risks, and Limitations

The dataset contains images of simulated falls by volunteers in controlled settings, therefore not capturing fully the complexity and unpredictability of real-life falls. Furthermore, the dataset may contain demographic bias, as the participants are all white.

BibTeX:

@article_Maldonado-Bascon_2019, author = Maldonado-Bascón, Sergio and Iglesias-Iglesias, Carlos and Martín-Martín, Pablo and Lafuente-Arroyo, Sergio, title = Fallen People Detection Capabilities Using Assistive Robot, place = Basel, publisher = MDPI, year = 2019, pages = 291 } APA:

Maldonado-Bascón, S., Iglesias-Iglesias, C., Martín-Martín, P., & Lafuente-Arroyo, S. (2019). Fallen people detection capabilities using assistive robot. Electronics.