Training Fusion Models

This page will provide a walk-through for training Fusion models in EdgeFirst Studio. For a walk-through on training Vision models, please see Training ModelPack.

Checkout our full video tutorial above as part of the EdgeFirst Studio Series to showcase the steps for running Fusion training in EdgeFirst Studio. Otherwise, follow along the steps shown below with section specific timestamps of the video.

Verify Dataset

Before running a training session, ensure the dataset is ready to be used for training.
This means that the dataset is properly annotated and the dataset is properly split with training and validation samples. The tutorial Verifying Datasets will show what to look for in a dataset before deploying it for training.

Select the Trainer Tool

Once the training dataset is ready, select "Train Experiments" from the tool options.

Tool Options

Specify the Project

Specify the project to run training at the center of the top menu bar.

Project Selection

Create Training Experiment

If you haven't already done so, create a training experiment. Create a new training experiment by clicking the "NEW EXPERIMENT" button on the top right.

Create New Experiment

This will provide pop-up for the user to specify the name and description of the experiment. Give a name and a description that reflects your goals in this experiment.

Create New Experiment

Click on the "CREATE NEW EXPERIMENT" button to create your new training experiment. This will show the created experiment.

Created Experiment

Open the created experiment by clicking on the experiment. Inside the experiment, we can create multiple training sessions. Each training session will train Fusion models which we will explore next.

Inside the Experiment

Create Training Session

Create a new training session within this experiment by clicking the "NEW SESSION" button as shown below.

Training Session

Configure the settings on the left panel by specifying "Trainer Type" to "EdgeFirst Fusion" and provide additional configurations for the name of the session and the dataset to deploy. Next configure the settings on the right panel by specifying training parameters. By default the Fusion model is configured with both camera and radar inputs, however, a Camera-Only or Radar-Only model are possible variations.

Additional information on these parameters are provided by hovering over the info button. For more information on available vision augmentations please see Vision Augmentations.

Training Options

Note

For an indoor setting, the "Radar Range Mode" is typically set to "Ultra Short (9m)" and the "Object Detection Range" is set to 9 meters. This is the maximum range of detection, further distances are ignored.

Start the Session

Start the session by clicking the "START SESSION" button on the bottom right.

Start the Session

Session Progress

The training session has now started while the progress is tracked on the left panel and additional information and status is shown on the right panel.

Training Session

The completed session will look as follows.

Completed Session

Training Session Attributes

Training Outcomes

Training Metrics

The training metrics are shown by clicking the button that views the training charts on the top right of the session card.

Training Metrics

Completed Session

Once completed, the status will be shown as complete.

Completed Session

Trained Models

The trained Keras, TFLite, and ONNX models are listed on the right. These models can be downloaded by clicking on the downward arrows on the right.

Session Details

Info

You can visualize the architecture of these models using https://netron.app/.

Next Steps

Now that you have generated your Fusion model, follow these next steps for validating your Fusion model.