Prototypical Networks with Random Weights for image classification on Omniglot and mini-ImageNet. Made with Jupyter.
Follow the instructions in the notebook and have fun!
It is made with Python3 and tested on Linux.
This project was based on:
- Cyprien Nielly implementation of Prototypical Networks.
- The original implementation, which can be found in Jake Snell's Github.
The idea of PNs can be originally found in Prototypical Networks For Few-shot Learning.
It's worth mentioning that using weights in order to calculate the prototypes is in Improved prototypical networks for few-Shot learning.