SketchRNNIntroToML
This is a project to replicate sketch rnn on mnist Credit goes to : https://github.com/magenta/magenta-demos/blob/master/jupyter-notebooks/Sketch_RNN.ipynb and https://github.com/MarioBonse/Sketch-rnn
Note:If You are unable to see the file in gradescope please view on github: https://github.com/nn210/SketchRNNIntroToML/blob/main/SketchRNN_Magenta_Final.ipynb
To View Our data preparation please refer: https://github.com/nn210/SketchRNNIntroToML/blob/main/Making_the_training_data.ipynb
This notebook is designed to build a neural network that generates sketch images based on the structure sefined by this paper: https://arxiv.org/pdf/1704.03477.pdf
The approach to train the sketch rnn on the MNIST dataset was inspired by the approaches given by: https://blog.otoro.net/2015/12/28/recurrent-net-dreams-up-fake-chinese-characters-in-vector-format-with-tensorflow/ and https://medium.com/analytics-vidhya/analyzing-sketches-around-the-world-with-sketch-rnn-c6cbe9b5ac80
Contributors :
Neeraja Narayanswamy
Sahil Makwane