You can use Amazon Lookout for Vision to identify missing components in products, damage to vehicles or structures, irregularities in production lines, minuscule defects in silicon wafers, and other similar problems. It uses machine learning (ML) to see and understand images from any camera as a person would, but with an even higher degree of accuracy and at a much larger scale.
Lookout for Vision allows customers to eliminate the need for costly and inconsistent manual inspection, while improving quality control, defect and damage assessment, and compliance. In minutes, you can begin using Lookout for Vision to automate inspection of images and objects–with no ML expertise required.
Dashboard view
The Amazon Lookout for Vision console provides a holistic view across all your production lines with an easy-to-use dashboard. The dashboard shows the projects by most defects, recent defects, and highest anomaly ratio, which enables you to quickly identify the production lines and processes that need immediate attention.
Simplified labeling
The Amazon Lookout for Vision console provides a visual interface to label your images quickly and simply, by applying a normal or anomaly label to the entire image with a click of a button. You can also select multiple images and apply the label to the entire selection with one click. If you organize your images in Amazon Simple Storage Service (Amazon S3) into Normal and Anomaly folders, you can use the Auto Labeler feature to automatically label each image when creating a dataset. If you have a large data set, you can also use Amazon SageMaker Ground Truth to efficiently label your images at scale.
Quick evaluation
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.
Trial anomaly detection tasks and feedback
You can instantly 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.
Using your trained models at the edge
You can use your trained Amazon Lookout for Vision models on a hardware device of your choice. Your trained models can be deployed on any NVIDIA Jetson edge appliance or x86 compute platform running Linux with an NVIDIA GPU accelerator. 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.
Manufacturing line integration
You can integrate Amazon Lookout for Vision with your manufacturing lines and implement automated visual inspection workflows for your use cases with just a few clicks in the Amazon Lookout for Vision console and a few API parameters. The Amazon Lookout for Vision API also enables you 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.
Get started building with Amazon Lookout for Vision in the AWS Management Console.