8000 GitHub - marcoslupion/THPoseLite: This repository incorporates THPoseLite, a neural network that obtains the pose from a person in thermal images.
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

This repository incorporates THPoseLite, a neural network that obtains the pose from a person in thermal images.

Notifications You must be signed in to change notification settings

marcoslupion/THPoseLite

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 

Repository files navigation

THPoseLite

This repository incorporates THPoseLite, a neural network that obtains the pose from a person in thermal images.

Usage

Run the script THPoseLite.py with the parameter --name indicating the neural network that will be created. Options are:

  • ResNet50
  • UNET
  • MobileNetV2

In the folder TrainedModels/ from Google Drive, pre-trained models with a dataset recorded in the University of Almería are incorporated. For each of the model, there are three configurations:

  • Tensorflow model. Weights are stored in 64-bit precision format.
  • TFLite model. Model that can be incorporated in smartphones and embedded devices.
  • Quantized model. Weights are stored in 8-bit precision format.

License

This software is provided 'as is', with the best of intentions but with any kind of warranty and responsibility of stability and correctness. It can be freely used, distributed and modified by anyone interested at it but a reference to this source is required.

Contact

For any question or suggestion feel free to contact:

About

This repository incorporates THPoseLite, a neural network that obtains the pose from a person in thermal images.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

0