OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Models

Quickdraw

Quickdraw is an open-source Python computer-vision project that includes a trained neural network for classifying sketches across 20 categories, a camera-based drawing app, and instructions for training with Quick, Draw! dataset files.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Quickdraw — OpenViST project profile in AI Models.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it to study a compact sketch-classification workflow or explore camera-based drawing with a colored object. The documented training path also supports changing the selected dataset classes.

Who it is for

  • People learning computer vision through sketch classification and object tracking
  • Developers exploring a small Python neural-network training project
  • Users interested in drawing by moving a colored object in front of a camera

Documented uses

  • Explore a trained classifier for sketches in 20 categories
  • Train a sketch classifier using Quick, Draw! dataset files
  • Use a camera to track a colored object and draw through its movement
Before you choose

Requirements & limits

The public GitHub repository is released under the MIT license.

  • The documented classifier covers 20 categories; adapting the class selection requires changing src/config.py and obtaining the matching dataset files.
  • The camera app expects a blue, red or green object and uses a laptop webcam by default.
  • The README mentions a canvas drawing app but says its script and demo were not yet released.
  • Listed requirements are Python 3.6, cv2, PyTorch and NumPy.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Train the classifier on selected Quick, Draw! categories

Prepare the documented dataset inputs and start model training for the chosen classes.

Before you start

  • Python 3.6
  • OpenCV (cv2), PyTorch and NumPy
  • Quick, Draw! dataset NPZ files for each selected category

Steps

  1. Download the NPZ files for the 20 documented classes, or for the categories selected for training.

  2. Store the files in the repository's data folder.

  3. For a different class selection, change the CLASSES constant in src/config.py and download the corresponding NPZ files.

  4. Run python3 train.py from the repository.

Expected outcome

The documented training process starts with the selected dataset classes; the README describes a 20-class trained model and reports an experiment using an 80:20 training/test split.

Guide limitations

  • The README does not specify training hardware, runtime or output-handling details.
  • The described experiment used the first 10,000 images for each class; this is not stated as a requirement for every training run.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
16 October 2026
Original website title
vietnh1009/QuickDraw · Official homepage title · retained in its source language
How we review sourcesSuggest a correction