gram-ai/capsule-networks
A PyTorch implementation of the CapsNet architecture for image classification using dynamic routing between capsules.
Not currently ranked — collecting fresh signals.
star history
This repository implements the CapsNet architecture from the 2017 paper by Sabour, Frosst, and Hinton. It trains a multi-layer capsule network on MNIST for digit recognition, using an iterative routing-by-agreement mechanism where lower-level capsules send outputs to higher-level capsules based on vector agreement. The implementation supports CUDA acceleration and uses TorchNet for training orchestration.
Frequently asked
- What is gram-ai/capsule-networks?
- A PyTorch implementation of the CapsNet architecture for image classification using dynamic routing between capsules.
- Is capsule-networks open source?
- Yes — gram-ai/capsule-networks is an open-source project tracked on heatdrop.
- What language is capsule-networks written in?
- gram-ai/capsule-networks is primarily written in Python.
- How popular is capsule-networks?
- gram-ai/capsule-networks has 1.8k stars on GitHub.
- Where can I find capsule-networks?
- gram-ai/capsule-networks is on GitHub at https://github.com/gram-ai/capsule-networks.