DeepCloud: Face Recognition Illustration
Last updated
Last updated
DeepCloud is the application platform of the Video Sense solution. Other tool in video sense incudes VMS, and AICraft, which is the tool used to generate AI models.
The system diagram of the video sense solution is shown below:
Resources of other tools part of Video Sense Solution:
The AIBox container can be configured to run various of AI video analytics algorithms, such as human detection, face detection, vehicle detection and license plate detection, etc. In this article, we will go cover what deepcloud (the application platform) can offer when the AIbox container has face detection algorithm running.
When the staff of retail store suspects shoplifting, security needs to review the playback footage tens of hours.
Shoplifting becomes a big cost for retail, and some store operators being to use shame wall to deter the potential shoplifter.
By digitize these shame wall, deepcloud can share these shoplifers across the entire retail chains. If one shoplifter steals in one store, and verified by the security, their faces can be stored and later the system can send the alert across the chain (user can define if he wants to share across all of his stores or just single store).
The alerts can be sent to the staff phone or dedicated terminal that resides near the checkout table or security.
Deepcloud web portal offers a convient way to check the faces and video footage.
When security identified a new theft, they can click the face to see all instances, and click “create a new subject” to add this face to the database. Again, this newly contributed theft will be shared across all stores (if authorized). When working with Video Sense VMS, clicking on the face will show the video clip of that moment.
Alerts will be pushed to mobile device and will be displayed under visiting records.
The cameras of all stores will be organized per store.
The face clustering works even during night:
