Train and deploy image classification models
You can create and deploy an image classification model onto your machine with Viam’s machine learning (ML) capabilities.
Manage the classification model fully on one platform: collect data, create a dataset and label it, and train the model for Single or Multi Label Classification.
Then, test if your model works for classifying objects in a camera stream or existing images with the mlmodel
classification model of vision service.
1. Collect Start by collecting images from your cameras with the data management service. You can view the data on the Data tab. |
---|
2. Create a dataset and label Once you have enough images of the objects you’d like to classify, label your data and create a dataset in preparation for training classification models. |
3. Train an ML model Use your labeled data to train your own models for object classification using data from the data management service. |
4. Deploy your ML model To make use of ML models with your machine, use the built-in ML model service to deploy and run the model. |
5. Configure an For object classification, you can use the vision service, which provides an ml model classifier model. |
6. Test your classifier Test your mlmodel classifier with existing images in the Viam app, live camera footage, or existing images on a computer. |
Next steps
After testing your classifier, see the following to further explore Viam’s data management and computer vision capabilities:
- Export Data Using the Viam CLI: Export your synced data from the Viam cloud.
- 2D Object Detection: Configure your machine’s camera to draw a bounding box around detected objects, based on a machine learning model.
- Update an existing ML model: Refine an existing ML model you have trained, and select which model version to deploy.
You can also explore our tutorials for more machine learning ideas:
Was this page helpful?
Glad to hear it! If you have any other feedback please let us know:
We're sorry about that. To help us improve, please tell us what we can do better:
Thank you!