clipx 0.3.0

Image background removal tool supporting U2Net and CascadePSP models

clipx

English | 简体中文

Introduction

clipx is an open-source Python library designed for quick and flexible image background removal. It supports both command-line interface (CLI) and Python API usage. Currently, it integrates two image segmentation models, U2Net and CascadePSP, which can be used individually or combined for enhanced performance. Additionally, clipx has an extensible architecture, making it easy to incorporate additional models in the future.

---

Installation

pip install clipx

---

CLI Examples

• Generate transparent image using combined mode (U2Net + CascadePSP):

clipx -i input.jpg

• Specify output path:

clipx -i input.jpg -o output.png

• Use fast mode for CascadePSP (faster but less accurate):

clipx -i input.jpg -o output.png --fast

• Use U2Net only for mask generation:

clipx -i input.jpg -o output.png -m u2net

• Enable debug logging:

clipx -i input.jpg --debug

---

Python API Example

# Import the background removal function
from clipx import remove_background

# Remove the background from an image with a single line of code
result = remove_background("photo.jpg")

print(f"Image with background removed saved to: {result}")

---

Advanced Usage

Logging Configuration

You can configure the logging level:

from clipx import set_log_level, enable_console_logging
import logging

# Enable console logging
enable_console_logging()

# Set log level
set_log_level(logging.DEBUG)

---

Acknowledgements and Code Sources

The image segmentation models in clipx are based on the following projects:

• U2Net: U^2-Net: Going Deeper with Nested U-Structure for Salient Object Detection
• CascadePSP: CascadePSP: Toward Class-Agnostic and Very High-Resolution Segmentation via Global and Local Refinement

We greatly appreciate the original authors' work and contributions.

---

License

This project is licensed under the MIT License.