# auto_object_removal

**Repository Path**: aizizai/auto_object_removal

## Basic Information

- **Project Name**: auto_object_removal
- **Description**: https://github.com/mintusf/auto_object_removal.git
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: main
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2024-03-05
- **Last Updated**: 2024-10-13

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

![Pytests](https://github.com/mintusf/auto-object-removal/actions/workflows/pytest.yml/badge.svg?branch=main)

# auto-object-removal
An application to automatically remove selected objects from images.
It combines image segmentation with image inpainting to create automated pipeline to remove whole objects.
There are two removal modes:
* all instance removal (semantic segmentation is used)
* single instance removal (instance segmentation is used)

Implemented segmentation models:
* Deeplab (torchvision weigths used)

Used image inpainting models:
* CR fill ([this repository](https://github.com/zengxianyu/crfill))

The project is WIP.

# How to use

1. Setup and enter pipenv environment
```bash
pipenv sync
pipenv shell
```

2. Run the program
   1. Semantic segmentation (all instances of selected classes are removed)

```bash
python app.py --mode semseg_img
```
Examples:

|  Image  | Video |
| ----------- | ----------- |
| ![](samples/GUI/sem_seg_GUI.PNG)      | ![](samples/GUI/sem_seg_GUI.gif)       |


   2. Instance segmentation (only selected instances are removed)

```bash
python app.py --mode instseg_img
```
Examples:

|  Image  | Video |
| ----------- | ----------- |
| ![](samples/GUI/inst_seg_GUI.PNG)      | ![](samples/GUI/inst_seg_GUI.gif)       |