Support for Binary Encoded Image Input Data¶
While OpenVINO models don’t have the ability to process images directly in their binary format, the model server can accept them and convert automatically from JPEG/PNG to OpenVINO friendly format using built-in OpenCV library. To take adventage of this feature, there are two requirements:
Model input, that receives binary encoded image, must have a proper shape and layout. Learn more about this requirement in input shape and layout considerations document.
Inference request sent to the server must have ceratain properties. These properties are different depending on the API (KServe or TensorFlow Serving) and interface (gRPC or REST). Choose the API you are using and learn more:
It’s worth noting that with KServe API, you can also send raw data with or without image encoding via REST API. This makes KServe REST API more performant choice comparing to json format in TFS API. The guide linked above explains how to work with both regular data in binary format as well as JPEG/PNG encoded images.