Amazon Lookout For Vision

Amazon Lookout for Vision is an ML service that uses computer vision to spot defects in manufactured products at scale.

The Amazon Lookout for Vision console provides a holistic view across your production lines with an easy to use dashboard. The dashboard shows the projects by most defects, recent defects, and highest anomaly ratio, which allows you to quickly identify the production lines and processes that need immediate attention.

The Amazon Lookout for Vision console provides a visual interface to label your images quickly and simply. The interface allows you to apply a normal or anomaly label to the entire image with a simple click of a button. You can also select multiple images and apply the label to the entire selection with one click. Alternatively, if you have a large data set, you can use Amazon SageMaker Ground Truth to efficiently label your images at scale.

Evaluate your anomaly detection model’s performance on your test dataset. If you do not provide your own test dataset, Amazon Lookout for Vision can automatically create a test dataset for you to evaluate your model’s performance. For every image in the test dataset, you can see a side-by-side comparison of the model’s prediction vs. the label assigned. You can also review detailed performance metrics such as precision/recall metrics, F1 score, and confidence scores.

You can run test detection tasks on additional images to get normal or anomaly predictions using your model. You can track your predictions, correct any mistakes, and provide the feedback to retrain newer models to improve anomaly detection accuracy.

You can use your trained Amazon Lookout for Vision models on various hardware devices. You can use AWS IoT Greengrass V2 to deploy, and manage your edge compatible customized models on your fleet of devices. AWS IoT Greengrass is an open source Internet of Things (IoT) edge runtime and cloud service that helps you build, deploy, and manage IoT applications on your devices.

You can deploy the same Amazon Lookout for Vision models that you've trained in the cloud onto AWS IoT Greengrass V2 compatible edge devices. You then use your deployed model to perform anomaly detection on premises without having to stream data continuously to the cloud. This allows you to minimize bandwidth costs and detect anomalies locally with real time image analysis.

Simply package your model as an AWS IoT Greengrass component from the Amazon Lookout for Vision console by choosing your target hardware device. Once your model is packaged as an AWS IoT Greengrass component you can directly deploy your model onto the AWS IoT core device of your choice.

You can integrate Amazon Lookout for Vision with your manufacturing lines and implement automated visual inspection workflows for your use cases with the Amazon Lookout for Vision console and a few API parameters. The Amazon Lookout for Vision API is designed to integrate your workflows into Amazon Augmented AI (Amazon A2I) for human review and verification by your process engineers so that they can continuously improve the accuracy of your models.

For additional information about service controls, security features and functionalities, including, as applicable, information about storing, retrieving, modifying, restricting, and deleting data, please see https://docs.aws.amazon.com/index.html. This additional information does not form part of the Documentation for purposes of the AWS Customer Agreement available at http://aws.amazon.com/agreement, or other agreement between you and AWS governing your use of AWS’s services.