![Secury360 logo](https://voxel51.com/wp-content/uploads/2023/04/Secury360-1.png)
![Secury360 logo](https://voxel51.com/wp-content/uploads/2023/04/Secury360-1.png)
Secury360 uses FiftyOne to balance computer vision datasets & improve model performance
The Secury360 box transforms CCTV setups into proactive perimeter detection solutions that use AI to eliminate false alarms and guarantee only human detection, with 99.998% accuracy. The Secury360 box features edge AI to gradually learn the terrain through deep learning to recognize behavior and discern if someone has bad intentions. Because only human detections get through the filter, operators have more time to respond to and prevent real threats.
![secury360 product hero image](https://voxel51.com/wp-content/uploads/2023/04/secury360-product-hero-1024x726.jpeg)
![secury360 product hero image](https://voxel51.com/wp-content/uploads/2023/04/secury360-product-hero-1024x726.jpeg)
Secury360's success with FiftyOne
Secury360 has to manage very large image and video datasets that are constantly being fed by devices. FiftyOne helps Secury360 constantly improve their machine learning model by enabling them to compare similar data, and in turn, deliver a more distributed dataset to train their surveillance model on.
Workflows & highlights
![multimedia icon images & videos multimedia icon](https://voxel51.com/wp-content/uploads/2023/04/IMAGE-VIDEO-DATASETS_T.png)
![multimedia icon images & videos multimedia icon](https://voxel51.com/wp-content/uploads/2023/04/IMAGE-VIDEO-DATASETS_T.png)
Image & video support
![BALANCED DATASETS icon balance datasets icon](https://voxel51.com/wp-content/uploads/2023/03/BALANCED-DATASETS_T.png)
![BALANCED DATASETS icon balance datasets icon](https://voxel51.com/wp-content/uploads/2023/03/BALANCED-DATASETS_T.png)
Balanced datasets
![model evaluation icon model evaluation icon - final](https://voxel51.com/wp-content/uploads/2023/03/MODELEVAL_T.png)
![model evaluation icon model evaluation icon - final](https://voxel51.com/wp-content/uploads/2023/03/MODELEVAL_T.png)