This project contains code and pre-trained sketch recognition models described in our ACMMM 2016 paper
This code is released under the MIT License (Please refer to the LICENSE file for details).
Please cite our paper if you happen to use our code and/or models:
If you use these models in your research, please cite:
@article{RK2015,
author = {Ravi Kiran Sarvadevabhatla and Jogendra Kundu and Venkatesh Babu R},
title = {Enabling My Robot To Play Pictionary : Recurrent Neural Networks For Sketch Recognition},
journal = {arXiv preprint arXiv:1608.03369},
year = {2016}
}
We provide two kinds of models -- CNN-based and RNN-based. Barring SketchCNN, the CNN-based models are Caffe-based. The RNN-based models have been obtained using the Lasagne framework. The code has been tested on Ubuntu 14.04 on the following NVIDIA GPUs: TITAN X, NVIDIA K40, NVIDIA K20.
To install Caffe, follow the instructions here
To install Lasagne, follow the instructions here
Refer to README.txt within the folders for details.
If you have a question or suggestion, please email jogendranathkundu@gmail.com and ravika@gmail.com