Aerial View Surveillance Dataset
Paper: Submitted
GitHub Link: https://github.com/iscaas/AFOSR-HAR-2021-2025/tree/main/OD-VIRAT (Dataset access link will be given in GitHub)
Description: OD-VIRAT is an object detection benchmark developed purely for object detection tasks in challenging surveillance environments. This dataset is constructed from videos of the VIRAT Ground 2.0 dataset, comprising 329 surveillance videos captured through stationary ground cameras mounted at significant heights (mostly atop buildings), spanning 11 distinct scenes. The recorded scenes include construction sites, parking lots, streets, and open outdoor areas. Different models of high definition (HD) cameras are used to capture scenes at different resolutions (1920×1080 and 1280×720) and frame rates (25∼30 frames per second (FPS)).
This dataset has two distinct versions that include OD-VIRAT-Large and OD-VIRAT-Tiny dataset. The statistical details of both datasets are given below in the table.
| Dataset | Number of Scenes | Number of Videos | Number of Images | Resolution | ||||||
| Train | Validation | Test | Train | Validation | Test | Train | Validation | Test | ||
| OD-VIRAT-Large | 10 | 10 | 8 | 156 | 52 | 52 | 377686 | 137971 | 84339 | (1920×1080), (1280×720) |
| OD-VIRAT-Tiny | 10 | 10 | 8 | 156 | 52 | 52 | 12501 | 4573 | 2786 | (1920×1080), (1280×720) |
Salient Soccer Events in Videos Dataset
Paper: https://ieeexplore.ieee.org/abstract/document/9530232
GitHub Link: https://github.com/iscaas/AFOSR-HAR-2021-2025/tree/main/SVE-Dataset (Dataset can be accessed through GitHub)
Description: The soccer video events (SVE) dataset comprises of short video clips of six different soccer events, including goal, penalty save, penalty goal, card, head goal, and substitute. The SVE dataset contains event videos captured from different views with both far and close field of views that offer great variety in the data. The SVE dataset is created in three distinct phases: 1) collection of soccer videos from multiple sources, such as UEFA Champions League, English Premier League, FIFA World Cup 2018, Bundesliga, and Primera Division; 2) extraction of event-specific short video clips from the downloaded soccer videos; and 3) annotation of the event-specific video clips with start and end boundary of the event. The SVE data set contains a total of 600 short video clips, which are divided into three subsets, including train, validation, and test sets with a split ratio of 60%, 20%, and 20%, respectively. The detailed information of dataset is presented in the table given below:
| Dataset | # of Video Clips | Video Extension | Clip Duration | FPS |
| Train | 360 | MP4 | 3 sec to 6 sec | 29 |
| Validation | 120 | MP4 | 3 sec to 6 sec | 29 |
| Test | 120 | MP4 | 3 sec to 6 sec | 29 |
Maize Nitrogen Stress Dataset
Paper: Coming soon
GitHub Link: Coming soon (Dataset access link will be given in GitHub)
Description: The maize nitrogen stress dataset captures data for monitoring nitrogen stress in maize during various developmental stages. This data set is collected from the Kansas State University North Farm Agronomy Education Center (AEC). The total area of the plot is approximately 1875 square feet, divided into 12 rows. The plot is further split into two halves: Nitrogen Surplus Rows and Nitrogen Deficient Rows. Each half is clearly marked to maintain an accurate record of the area with specific nitrogen concentrations.
The data set includes two types of images: RGB images captured with a Canon SX530 HS 16MP camera and multi-spectral images captured with a Micasense RedEdge MX camera, covering wavelengths from 400 nm to 900 nm.