Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/shvbsle/sputils

Utils and Notebooks for building deep-learing model with superpaper
https://github.com/shvbsle/sputils

dataset deep-learning generative-art sketch-rnn

Last synced: about 14 hours ago
JSON representation

Utils and Notebooks for building deep-learing model with superpaper

Awesome Lists containing this project

README

        

# Superpaper Utils
Utils and Notebooks for building deep-learing models with data-generated from [superpaper web-app](https://superpaper.netlify.app/).

P.S: Most images in dataset are gonna be low in quality as thats how good my art-skills are

# Data

The stroke-data is exported as a json file. Use the ExploreStrokes.ipynb notebook to use utils for exploring data.

Currently, a single data-point comprises on 3 elements:

1) Stroke JSON
2) Base image
3) Sketch Layer

---

## 1. Stroke JSON
```
{
"description": "",
"stroke": [{
"layer": "",
"type": "",
"memento": [
[, , , ],
[809, 124, -310, -10],
.
.
[]
]
}, {
"layer": "Layer 1",
"type": "mouseup",
"memento": [
[811, 126, -310, -10],
[811, 126, -310, -10],
.
.
.
[]
]
}
.
.
.],
"device_type": "pc",
"canvas_h": 649,
"canvas_w": 999
}
```

Note that an empty entry in memento (i.e. []) signifies end of stroke

## 2. Base Image
This is the image used as reference for drawing the data

![alt text]()

## 3. Sketch Layer
This is the image that was exported after drawing (p.s forgive me for this garbage trace)

![alt text]()

---

# TODO
- [ ] Add Sketch-RNN wrapper
- [X] Create section for comic