-
Book Overview & Buying
-
Table Of Contents
-
Feedback & Rating

Computer Vision on AWS
By :

Throughout this book, we have discussed many CV use cases and provided steps to deploy models on AWS for inference. What if these models are unable to be deployed to the cloud? To perform model training, we often deal with vast amounts of data and large-sized images. In some instances, you may encounter legal restrictions or security regulations that require this data to be processed on-premises. You may also be processing video from a camera on a manufacturing floor and discover that there are bandwidth constraints that limit how much video content is sent to the cloud. In healthcare, medical images that are generated by CT scans and MRI machines provide crucial information for clinicians and providers. These images need to be processed in near real time for quick delivery of the right treatment plan to a patient. Uploading the images to the cloud would incur high latency and delay the ML model insights used to help make clinical decisions. Running ML...