https://github.com/nikhilroxtomar/unet-for-person-segmentation

U-Net for person segmentation in TensorFlow using Keras API.
https://github.com/nikhilroxtomar/unet-for-person-segmentation

keras tensorflow unet-image-segmentation unet-keras unet-segmentation

Last synced: about 1 year ago
JSON representation

U-Net for person segmentation in TensorFlow using Keras API.

• Host: GitHub
• URL: https://github.com/nikhilroxtomar/unet-for-person-segmentation
• Owner: nikhilroxtomar
• Created: 2021-01-12T12:18:07.000Z (over 5 years ago)
• Default Branch: main
• Last Pushed: 2021-05-27T09:36:48.000Z (about 5 years ago)
• Last Synced: 2025-03-28T08:11:19.849Z (over 1 year ago)
• Topics: keras, tensorflow, unet-image-segmentation, unet-keras, unet-segmentation
• Language: Python
• Homepage:
• Size: 21.8 MB
• Stars: 25
• Watchers: 1
• Forks: 12
• Open Issues: 2
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# Unet for Person Segmentation

We are using the famous UNet architecture for segmenting person from an image. For the person segmentation, we are going to use the person segmentation dataset. U-Net is built for Biomedical Image Segmentation. It is the base model for any segmentation task. It follows an encoder-decoder approach. It used skip connection to get the local information during downsampling path and use it during the upsampling path. 

YouTube Video: [https://youtu.be/qrL22HEaUGA](https://youtu.be/qrL22HEaUGA) 


Arxiv Paper: [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://arxiv.org/abs/1505.04597])

These images are generated after the model is trained on 2 epochs.

![Person Segmentation](save_images/Black-Widow-Avengers.jpg)

![Person Segmentation](save_images/evolution-of-thor-5.png)

![Person Segmentation](save_images/Robert-Downey-Jr-3-640x433.jpg)