In this video, we will learn to segment human images using a DeepLabV3+ architecture enhanced with a channel-wise attention mechanism (squeeze and excitation network). Here we have used the Person Segmentation dataset for which consists of images of humans with their respective annotated masks. All the code is written in Python programming language in TensorFlow 2.5.0 framework.

In this video, we will cover the entire process for human image segmentation, which include:

1. Data processing

2. Network architecture

3. Training

4. Evaluation

DeepLabv3+ is a semantic segmentation architecture that improves upon DeepLabv3 with several improvements, such as adding a simple yet effective decoder module to refine the segmentation results.

The Squeeze-and-Excitation Block is an architectural unit designed to improve the representational power of a network by enabling it to perform dynamic channel-wise feature recalibration

Code (GitHub): https://github.com/nikhilroxtomar/Hum…

Person Segmentation Dataset: https://www.kaggle.com/nikhilroxtomar…

Subscribe: https://www.youtube.com/c/IdiotDeveloper/featured

#tensorflow #deeplabv3