Skip to content

Instance Segmentation Mask R-CNN

The Instance Segmentation Mask R-CNN model provides state of the art Instance segmentation and instance segmentation on images with fast inference times.

Instance Segmentation Example

Launch on SageMaker.


  • State of the art accuracy on the COCO 2017 dataset
  • Box AP: 43.0
  • Based on the Mask R-CNN paper
  • Network Architecture: ResNet + FPN (Feature Pyramid Network): ResNeXt - 101
  • Flexible endpoint to make inferences on images: Returns a segmented image in different formats or the metadata as JSON
  • Only pay for what you use with a simple metered pricing model
  • Same price independent of the resources (memory/cpu/gpu) used


This model has a fee of 0.15 dollars per hour for real time inference and 1 dollar per hour for batch jobs.

You will incur costs for software use only for as long as the Endpoint is running.

AWS infrastructure costs are independent and in addition to the costs of the software and it depends on the instance type selected to host the algorithm.