Updating readme

Author: emreaksan

README.md

@@ -48,4 +48,22 @@ where `UNIQUE_ID` is the same as above.
 You can download [our main model](https://drive.google.com/drive/folders/1C6m7dbXaL4wn5Z4-K7ZniqoZaNTiQBdP?usp=sharing) and run evaluation script as explained above.
 
 ### Demo
-An interactive demo is provided in the [smarting-js](https://github.com/eth-ait/cose/tree/master/smartink-js) folder, using the pre-trained models shared above.   
+An interactive demo is provided in the [smarting-js](https://github.com/eth-ait/cose/tree/master/smartink-js) folder, using the pre-trained models shared above.
+
+### Citation
+```
+@article{aksan2020cose,
+  title={CoSE: Compositional Stroke Embeddings},
+  author={Aksan, Emre and Deselaers, Thomas and Tagliasacchi, Andrea and Hilliges, Otmar},
+  journal={Advances in Neural Information Processing Systems},
+  volume={33},
+  year={2020}
+}
+
+@article{gervais2020didi,
+  title={The DIDI dataset: Digital Ink Diagram data},
+  author={Gervais, Philippe and Deselaers, Thomas and Aksan, Emre and Hilliges, Otmar},
+  journal={arXiv preprint arXiv:2002.09303},
+  year={2020}
+}
+```   

smartink-js/README.md

@@ -38,4 +38,30 @@ You still need to implement a model class in JS to load the pre-trained models a
 The user interface is implemented in the `reconstruct.js` and `predict.js` by modifying David Ha's [demo](https://github.com/hardmaru/sketch-rnn-flowchart).
 
 
-Finally, you can store your TF-JS models on storage.googleapis without requiring a dedicated server.     
+Finally, you can store your TF-JS models on storage.googleapis without requiring a dedicated server.     
+
+### Citation
+If you are using this demo in your work, please cite the following:
+```
+@article{aksan2020cose,
+  title={CoSE: Compositional Stroke Embeddings},
+  author={Aksan, Emre and Deselaers, Thomas and Tagliasacchi, Andrea and Hilliges, Otmar},
+  journal={Advances in Neural Information Processing Systems},
+  volume={33},
+  year={2020}
+}
+
+@article{gervais2020didi,
+  title={The DIDI dataset: Digital Ink Diagram data},
+  author={Gervais, Philippe and Deselaers, Thomas and Aksan, Emre and Hilliges, Otmar},
+  journal={arXiv preprint arXiv:2002.09303},
+  year={2020}
+}
+
+@article{ha2017sketchrnn,
+  title={A neural representation of sketch drawings},
+  author={Ha, David and Eck, Douglas},
+  journal={arXiv preprint arXiv:1704.03477},
+  year={2017}
+}
+```