Overview:

TIMo (Time-of-Flight Indoor Monitoring) is a dataset of infrared and depth videos intended for the use in Anomaly Detection and Person Detection/People Counting. It features more than 1,500 sequences for anomaly detection, which sum up to more than 500,000 individual frames. For person detection the dataset contains more than than 240 sequences. The data was captured using a Microsoft Azure Kinect RGB-D camera (→ specs). In addition, we provide annotations of anomalous frame ranges for use with anomaly detection and bounding boxes and segmentation masks for use with person detection.

The data was captured in parts from a tilted view and a top-down perspective. See below for examples of how the data looks like.

More details on the annotations and the data format can be found here and in our paper.

Download:

The dataset is available for download at this page (Note: You will have to create a free account to download the data).

The whole dataset has a size of roughly 110 GiB split into an anomaly detection dataset and one for person detection/people counting.

Citation and Licensing Terms:

Please make sure to read about the licensing terms here in case you want to use the dataset.

Example Frames:

Infrared

Depth
Infrared
Depth

Demo Video:

Recording Setups:

Data was recorded using Microsoft Azure Kinect for in the tilted- as well as the top-down view setup.