Release Notes
SmartVis Face Server
The link below gives the full installation manual for the server.
The recommended installation is a native installation onto a server running Ubuntu 18.04 x64. This will use your NVIDIA GPU and is required for processing large amounts of image and video data.
Please note, after installing this you will need to download the server images from Dockerhub. Please contact Digital Barriers support with your validated User ID so that we can give you access to the server images.
You can also download SmartVis Face Server as a Virtual Disk Image (VDI). This can be imported into most Virtual Machine Managers, such as Virtualbox.
SmartVis Face Server Virtual Disk Image (9.88 GByte)
This type of installation is fine if you are using SmartVis Face Server for synchronising with Identifier and processing low volumes of data. If you are uploading video, it is recommended to use a native installation.
Once installed, the SmartVis Face Server can be configured and used via the User and Admin Web Portals.
Optional packages can be installed for enhanced security.
Mobile Clients
The SmartVis Face Server can also be accessed via a compatible Android Mobile Phone.
If you wish to use iOS please contact Digital Barriers support. We will need the email registered with your Apple account.
Bodyworn Clients
The following manual gives details on how to set-up your Bodyworn devices to use SmartVis Face Server.
EdgeVis and Secure Connect
To configure SmartVis Face Mobile to synchronise the watchlists and alarms over EdgeVis.
Batch Enrollment
If you have a lot of image and video data, complete with subject meta-data, you can batch import into SmartVis Face.
Hardware specification for Server
For a native installation:
The ONLY supported hardware for this type of installation is the Dell R740xd, dedicated to SmartVis Face Server running Ubuntu 18.04:
Single or Dual Processor
Intel Xeon Silver 4114 @ 2.20 GHz
10 cores
CPU Benchmark of 14,360
1 to 3 Tesla T4
Depends on use-case and number of streams: 2 x 4TB disks
Setup in RAID 1 (mirroring): 64 GByte RAM
If installing a virtual image:
The SmartVis Face Engine is supplied as a Virtualbox VDI image, ready to be installed into your virtualised environment. In the manual we document how to set-up an instance using Virtualbox. However, it is possible to import VDI images into other Virtual Machine environments such as VMWare and KVM.
When creating the instance of the SmartVis Face appliance on your host you will need the following spare compute capacity to designate to the appliance:
A minimum of 4 cores with a minimum speed of 2.2GHz
8 Gigabytes of RAM
128 Gigabytes of storage
Note, these are the minimum requirements. Depending on the level of usage it may be desirable to allocate more resource to your instance. In particular, increasing the number of cores can significantly increase the computation speed. We recommend a base operating system of Ubuntu 18.04 x64 as the host operating system.