Skip to content

Detect what is in an image from an interactive app using pretrained models.

Notifications You must be signed in to change notification settings

mtanco/wave-grad-cam

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Image Detection with Keras and Wave

Using open source Keras and Grad-CAM the user can upload their own photos and get explainable predictions of what a standard, pre-trained model thinks is in the image. This app makes these open source tools more accessible by adding a easy-to-use front end.

This app can be used as a template for Wave App developers on how to integrate third-party modeling frameworks into your own apps.

Screen1 Screen2 Screen3

Local Development

Create a python environment in the home app directory and install requirements. Python 3.7 is recommended but not required.

python3 -m venv venv
./venv/bin/pip install -r requirements.txt

When running this application locally, ensure that you are using the parameter for pointing to the images/ folder so that these files will be automatically available in the Wave serve.

H2O_WAVE_PUBLIC_DIR="/images/@/Users/mtanco/h2o_wave/apps/wave-grad-cam/images" ./venv/bin/wave run src.app

Visit the app at http://localhost:10101

Publish to the AI App Store

After configuring your h2o CLI to a specific environment, you can upload this app using

h2o bundle import

About

Detect what is in an image from an interactive app using pretrained models.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages