https://github.com/timothygebhard/pytorch-conv4d

Rudimentary Conv4D Layer Implementation for PyTorch
https://github.com/timothygebhard/pytorch-conv4d

Last synced: 4 months ago
JSON representation

Rudimentary Conv4D Layer Implementation for PyTorch

• Host: GitHub
• URL: https://github.com/timothygebhard/pytorch-conv4d
• Owner: timothygebhard
• License: mit
• Created: 2019-01-07T15:19:35.000Z (over 5 years ago)
• Default Branch: master
• Last Pushed: 2019-01-07T15:40:04.000Z (over 5 years ago)
• Last Synced: 2024-01-16T10:46:29.110Z (5 months ago)
• Language: Python
• Size: 4.88 KB
• Stars: 36
• Watchers: 4
• Forks: 13
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Lists

• awesome-stars - timothygebhard/pytorch-conv4d - Rudimentary Conv4D Layer Implementation for PyTorch (Python)

README

        
# pytorch-conv4d

This repository contains a simple PyTorch port of the [conv4d for TensorFlow repository](https://github.com/funkey/conv4d) by Jan Funke. It consists essentially of a single class, `Conv4d`, which provides a (still rather rudimentary) PyTorch layer for 4-dimensional convolutions. Like the original, it works by performing and stacking several 3D convolutions (see the original repository for a more detailed explanations).

This implementation is still work in progress (hence it comes with no warranties whatsoever), and pull requests or advice for improvements are very much welcome! :)