# torch-dwt

**Repository Path**: Heconnor/torch-dwt

## Basic Information

- **Project Name**: torch-dwt
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2025-04-01
- **Last Updated**: 2025-04-02

## Categories & Tags

**Categories**: Uncategorized

**Tags**: Algorithm

## README

# 3D Discrete Wavelet Transform (DWT) im Pytorch

This package implements the 1D,2D,3D [Discrete Wavelet Transform](https://en.wikipedia.org/wiki/Discrete_wavelet_transform) and inverse DWT (IDWT) in Pytorch.
The package was heavily inspired by [pytorch_wavelets](https://github.com/fbcotter/pytorch_wavelets) and extends its functionality into the third dimension.

The wavelets are provided by the [PyWavelets](https://github.com/PyWavelets/pywt) package.

All operations in this package are fully differentiable. 

## Installation

Python > 3.9

```
git clone https://github.com/KeKsBoTer/torch-dwt
cd torch-dwt
pip install -e .
```

### distribution
```
# source distribute
python setup.py sdist

# binary distribute
python setup.py bdist

# wheel distribute
python setup.py bdist_wheel
```


## Example Usage

### Internal
- 3D

```python
from torch_dwt.functional import dwt3,idwt3
import torch

# 8 images with 3 color channels and size of 100x100
x = torch.rand(8,3,100,100,100)
coefs = dwt3(x,"sym2") # coefs of shape (1,2,3,50)
# reconstruct signal from coefficients
y = idwt3(coefs,"sym2")
```

- 2D

```python
from torch_dwt.functional import dwt2,idwt2
import torch

# 8 images with 3 color channels and size of 100x100
x = torch.rand(8,3,100,100)
coefs = dwt2(x,"db2") # coefs of shape (1,2,3,50)
# reconstruct signal from coefficients
y = idwt2(coefs,"db2")
```

- 1D

```python
from torch_dwt.functional import dwt,idwt
import torch

# batch of size 8 with 3 channels
x = torch.rand(8,3,100)
coefs = dwt(x,"haar") # coefs of shape (1,2,3,50)
# reconstruct signal from coefficients
y = idwt(coefs,"haar")
```

## Testing

For testing we compare our implementation againts [PyWavelets](https://github.com/PyWavelets/pywt).
This command runs the tests:
```bash
# navigate into torch-dwt directory
pytest .
```

## Changed

### 2025-04-02
- implement as a Pytorch layer, 1D/2D/3D