The *INFRADEL* dataset (INFRAred Dataset for Occupancy Estimation and Localization), is designed to facilitate research in occupancy estimation and localization using thermal imaging technology. This dataset contains thermal images, including 2-channel gray-scale thermal images for the ABFA* model and converted RGB format for the YOLO model. It captures a diverse range of scenarios, including both static and dynamic settings with varying levels of occupancy, making it ideal for developing and testing occupancy estimation algorithms.
Dataset Details:
- `Data Collection`: The dataset was collected over nine days at our Institute. It includes both static scenarios (seven days) and dynamic scenarios (two days) from different locations.
- `Locations`: Data was collected from two main locations at IISER Bhopal Institute:
- `Classroom`: Representing a dense occupant environment with up to 15 occupants.
- `Research Lab and Office`: Emulating a sparse occupant setting with a maximum of 8 occupants.
- `Occupancy Levels`: The dataset covers occupancies ranging from 0 to 15, where '0' indicates background-only images without any occupants.
Dataset Structure:
- The dataset is organized into two main categories:
- `Static`: Data collected from dense and sparse scenarios.
- `Dynamic`: Data collected from dynamic scenarios in office settings.
ABFA* Model Data
- `Path`: /ABFA*/DATA/
- `Structure`: Divided into Static and Dynamic folders containing thermal images for training.
- `Format`: Images are stored in .npy format with corresponding CSV files containing timestamps.
YOLO Model Data
- `Path`: /YOLO/DATA/
- `Structure`: Divided into Static and Dynamic folders containing RGB images used for training.
- `Labels`: Stored in YOLO .txt format with bounding box information for each occupant.
Sahoo, S. R., & Lone, H. R. (2024). Mapping Thermal Footprints: Occupancy Estimation and Localization in Diverse Indoor Settings with Thermal Arrays. ACM Conference of Computing and Sustainable Societies (ACM COMPASS 2024).
Contact: If you have any questions or need further clarification about the dataset, feel free to ask.
