Gastroenterology & Artificial Intelligence: 3rd Annual Global Artificial Intelligence Summit | September 2021-2-Computer Vision in Medicine

2-Computer Vision in Medicine_Bhotika

Pdf Summary

The document discusses the use of computer vision and health artificial intelligence (AI) at Amazon Web Services (AWS). It highlights the various AWS services in AI and machine learning, with a focus on computer vision and health AI services. The document also explains how models can be built for specific tasks by domain experts, even without extensive ML expertise.<br /><br />The AWS ML stack is described as a comprehensive set of machine learning capabilities. It includes AI services, ML services, frameworks, and infrastructure, all aimed at facilitating the development and deployment of machine learning models.<br /><br />Specific services mentioned in the document include Amazon Rekognition, Amazon Textract, Amazon Lookout for Vision, and Amazon Comprehend Medical. These services offer capabilities such as text extraction, table extraction, form extraction, celebrity recognition, face detection and analysis, and spot product defects using computer vision.<br /><br />The document emphasizes the importance of addressing the long tail of computer vision problems, which are often specialized and require minimal training data. It mentions the use of Amazon Rekognition Custom Labels, which enables the creation of models with tens to hundreds of images instead of tens of thousands, without requiring ML expertise.<br /><br />The document also discusses the use of interactive feedback and labeled and unlabeled datasets to improve models. It highlights the benefits of reducing labeling effort and providing feedback to improve model performance.<br /><br />Additional topics covered in the document include multimodal classification, spotting product defects using computer vision, and mining unstructured data in healthcare. It also mentions the democratization of AI and the need for ML methods that support semi-supervised learning, feedback, and learning from a few examples.<br /><br />In summary, the document provides an overview of computer vision and health AI services offered by AWS and discusses the use of AI and ML methods for building models and extracting knowledge from various types of data.

Keywords

computer vision

health artificial intelligence

Amazon Web Services

AWS

AI services

machine learning

computer vision services

Amazon Rekognition

Amazon Textract

Amazon Lookout for Vision