Convert Tensorflow Pretrained Model to IR with OpenVINO and Real time Video Prediction || AI Sangam

preview_player
Показать описание
This video deals with optimizing and deploying a model that was trained with the TensorFlow using OpenVINO Toolkit. We will download the trained tensorflow model from tensorflow zoo and convert it into optimized Intermediate Representation (IR). Again we will test the model in the Intermediate Representation format using the Inference Engine.

Steps for optimizing and deploying a model that was trained with the TensorFlow

Step1:

Step 2:

Convert a TensorFlow model to produce an optimized Intermediate Representation (IR) which produces both .bin and .xml file. For this step please run the below command

Step 3:

Step 4:

Please move to the folder using

cd /opt/intel/computer_vision_sdk_2018.5.455/deployment_tools/inference_engine/samples/python_samples/

Please run the file and use the video you want

I hope you may like the video. We will do continuous work in OpenVINO Toolkit and will put all efforts to bring high end deep learning models to run on lower end devices.

Please spare some time to visit us at:
Рекомендации по теме
Комментарии
Автор

Hi, Nice work!
Could you guide on how to run the same using MYRIAD devices?

balajichetty