Read output from a model asynchronously | Inference Engine | 2.3.0