Platforms And Applications

This VVAS release supports Zynq MPSoc based Embedded platforms like KV260 SOM and zcu104 platforms. This release has also enabled the support for Versal device based platforms. This release has been validated on zcu104 based platform for different usecases. Various supported example applications using VVAS on zcu104 and KV260 SOM based platforms are listed below. For creating example designs for Versal based platforms, you may refer to the zcu104 based example designs provided with this release as reference for patches needed for other platforms.

SOM Examples

There are three Application specific SOM platforms supported. Click on the link below to know more about these.

Note

Currently KV260 SOM example designs mentioned below may be based on VVAS 1.0 and Vitis AI 1.4 release. These designs will be upgraded to VVAS 2.0 and Vitis AI 2.5 soon.

Smart Camera

Smart Camera Application with face detection and display functionality. For more details refer to Smart Camera

AIBox-ReiD

Application demonstrating multistream tracking and Re-Identification. For more details, refer to AIBox-ReID

Defect Detection

Application to detect defects in mangoes. For more details, refer to Defect Detect

ZCU104 Examples

Smart Model Select

This application allows selection of ML model from 16 supported models, input source and output using command line and then it constructs and executes the pipeline. For more details, refer to Smart ML Model Select.

Multi Channel ML

This application demonstrates multi channel multi model Machine Learning capability using VVAS. Multi Channel ML Tutorial tutorial for beginners explains step by step approach to build this application/pipeline using VVAS.