tf-bodypix 0.1.0

Python implemention of the TensorFlow BodyPix model.

TensorFlow BodyPix (TF BodyPix)

A Python implementation of body-pix.

Install

Install with all dependencies:

pip install tf-bodypix[all]

Install with minimal or no dependencies:

pip install tf-bodypix

Extras are provided to make it easier to provide or exclude dependencies when using this project as a library:

extra name	description
tf	TensorFlow (required). But you may use your own build.
tfjs	TensorFlow JS Model support
image	Image loading via Pillow, required by the CLI.
all	All of the libraries

CLI

Creating a simple body mask

TF_CPP_MIN_LOG_LEVEL=3 \
python -m tf_bodypix \
    image-to-mask \
    --image /path/to/input-image.jpg \
    --output-mask /path/to/output-mask.jpg \
    --threshold=0.75

Colorize the body mask depending on the body part

TF_CPP_MIN_LOG_LEVEL=3 \
python -m tf_bodypix \
    image-to-mask \
    --image /path/to/input-image.jpg \
    --output-mask /path/to/output-colored-mask.jpg \
    --threshold=0.75 \
    --colored

Additionally select the body parts

TF_CPP_MIN_LOG_LEVEL=3 \
python -m tf_bodypix \
    image-to-mask \
    --image /path/to/input-image.jpg \
    --output-mask /path/to/output-colored-mask.jpg \
    --threshold=0.75 \
    --parts left_face right_face \
    --colored

API

import tensorflow as tf
from tf_bodypix.api import download_model, load_model, BodyPixModelPaths

bodypix_model = load_model(download_model(
    BodyPixModelPaths.MOBILENET_FLOAT_50_STRIDE_16
))

image = tf.keras.preprocessing.image.load_img(
    '/path/to/input-image.jpg'
)
image_array = tf.keras.preprocessing.image.img_to_array(image)
result = bodypix_model.predict_single(image_array)
mask = result.get_mask(threshold=0.75)
tf.keras.preprocessing.image.save_img(
    '/path/to/output-mask.jpg',
    mask
)

colored_mask = result.get_colored_mask(mask)
tf.keras.preprocessing.image.save_img(
    '/path/to/output-colored-mask.jpg',
    colored_mask
)

Acknowledgements

• Original TensorFlow JS Implementation of BodyPix
• Linux-Fake-Background-Webcam, an implementation of the blog post describing using the TensorFlow JS implementation with Python via a Socket API.
• tfjs-to-tf for providing an easy way to convert TensorFlow JS models
• virtual_webcam_background for a great pure Python implementation