Intel® Distribution of OpenVINO™ Toolkit
Community assistance about the Intel® Distribution of OpenVINO™ toolkit, OpenCV, and all aspects of computer vision-related on Intel® platforms.

Creating a new class for the model zoo

SEEDLE__RANDY
Beginner
622 Views

How could I go about adding a new class to the model zoo. I have looked at the model zoo and there really is nothing that would identify what I am looking for.

1. Caves, lava tubes, crevices.

2. Various mineral types.

3. Ice in various scenarios.

Is there a tool to do this ?

Randy

 

 

0 Kudos
3 Replies
Zulkifli_Intel
Moderator
600 Views

Hello Randy Seedle,

 

Thank you for reaching out to us.

 

If you cannot find the appropriate pre-trained model from Intel Open Model Zoo to run the inference, you can use your framework of choice to prepare and train a deep learning model. Training Custom Object Detector shows how to train a model in the TensorFlow framework.

 

Once you are done with the model training, you can use the Model Optimizer to converts the trained neural network from its source framework to an open-source, nGraph-compatible Intermediate Representation (IR) for use in inference operations. The Model Optimizer imports models trained in popular frameworks such as Caffe*, TensorFlow*, MXNet*, Kaldi*, and ONNX* and performs a few optimizations to remove excess layers and group operations when possible into simpler, faster graphs.

 

Sincerely,

Zulkifli


0 Kudos
Munesh_Intel
Moderator
568 Views

Hi Randy,

This thread will no longer be monitored since we have provided a solution. If you need any additional information from Intel, please submit a new question.



Regards,

Munesh


0 Kudos
Zulkifli_Intel
Moderator
542 Views

Hello Randy Seedle,

 

Here is an additional option you can use to train your model. You can use the OpenVINO™ Training Extensions to train your custom detection model, it provides a convenient environment to train Deep Learning models and convert them using the OpenVINO™ toolkit for optimized inference. Note that this is for limited models only.

 

Sincerely,

Zulkifli

0 Kudos
Reply