Yes, Quividi has 2 specific versions for ARM processors, such as those from the Cortex family.
Quividi also proposes an Android version for ARM. Because Android does not support external cameras, this version is mostly meant for tablets. Also, because Android does not let an application run in the background, this version comes as an SDK, to be integrated by developers into apps for tablets.
Note that ARM processors are generally delivering limited CPU, which Quividi makes use of, so that the use cases limit the use of these versions to situations where faces need to be within 5 meters (15 feet) of the camera.
The system requirement for our ARM versions to run in standard video resolution (640*480 delivering the above mentioned face detection distance) is 12,000 DMIPs over 2 cores – along with the support for the NEON instruction set.
For Android versions, and for a lesser face detection distance (320*240 delivering a 2.5 meter or 8 feet detection distance – typically adapted to small screens) the requirement is 4,000 DMIPs. The version of VidiReports for those platforms is called VidiReports Nano.
http://en.wikipedia.org/wiki/List_of_ARM_microarchitectures provides the elements to figure out the total DMIPs delivered by an ARM processor.