https://github.com/csinva/scattering-transform-experiments

Repository for experiments with scattering transforms
https://github.com/csinva/scattering-transform-experiments

ai convolutional-neural-networks deep-learning geometric-deep-learning machine-learning ml scale scattering-transform statistics

Last synced: 20 days ago
JSON representation

Repository for experiments with scattering transforms

• Host: GitHub
• URL: https://github.com/csinva/scattering-transform-experiments
• Owner: csinva
• Created: 2017-09-16T06:45:55.000Z (over 7 years ago)
• Default Branch: master
• Last Pushed: 2019-01-25T22:19:46.000Z (almost 6 years ago)
• Last Synced: 2024-10-28T12:02:34.439Z (2 months ago)
• Topics: ai, convolutional-neural-networks, deep-learning, geometric-deep-learning, machine-learning, ml, scale, scattering-transform, statistics
• Language: Jupyter Notebook
• Size: 44.3 MB
• Stars: 2
• Watchers: 5
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# scattering transform

- *code for development of adaptive scattering transform that works with conv on small data*

# setup

- download cifar data by running download.py in data folder

# references

- *note*: actual weights have to be downloaded to be used

- uses AlexNet code from: https://github.com/guerzh/tf_weights

- uses scattering transform code from: https://github.com/tdeboissiere/DeepLearningImplementations

# usage

parameters - M, N, J, L=8 (hidden in filters_bank function)

- M, N - size of images

- J - number of scales (window size at any scale j: )

- L - number of orientations