DeepFit
DeepFit is an open source package for creating novel methods that help all the stake holders better manage Patient Engagement. It leverages research technology like Data Shapley, Multi-Accuracy and cPCA from Stanford Artificial Intelligence Labs (SAIL)
Getting Started
Please Install DeepFit package using
!pip install deepfitv
Programming Guide
Incision Object Declaration
deepfitv.incision.Incision(<path of image>)
- To identify incision image object detection run the object detection function run
deepfitv.incision.Incision(<path of image>).object_detection()
- To classify the incisin image into less than 30 or post 30 days of surgery run
deepfitv.incision.Incision(<path of image>).classify_image()
- To use SinGAN feature, use a single image and generate random sample images from it refer :
deepfitv.SinGAN.help()
Prerequisites
deepfitv.incision module requires 'imageai' and 'tensorflow' libraries as dependencies
!pip install tensorflow
!pip install imageai
To run the Incision Object Image use case one would need to download the model h5 files which we have developed and copy in the directory
Incision/models
[https://github.com/Virtusa-vLife/DeepFit/releases/download/deepfit/Image_classification.h5]
[https://github.com/Virtusa-vLife/DeepFit/releases/download/deepfit/detection_model.h5]
[https://github.com/OlafenwaMoses/ImageAI/releases/download/essential-v4/pretrained-yolov3.h5]
download reference json at
Incision/json
[https://github.com/Virtusa-vLife/DeepFit/releases/download/deepfit/detection_config.json]
License
Coming Soon