Akin: Generating UI Wireframes From UI Design Patterns Using Deep Learning

Akin is a UI wireframe generator that allows designers to chose a UI design pattern and provides them with multiple UI wireframes for a given UI design pattern. Akin uses a fine-tuned Self-Attention Generative Adversarial Network trained with 500 UI wireframes of 5 android UI design patterns

---

Dataset

Akin uses a manually annotated subset of RICO dataset

• Dataset with annotations and semantic images

---

Setup and usage

Akin uses Python 3.7 and Tensorflow 2.1.

To install and retrain Akin, follow the steps below

• Download and extract the dataset to the data/train directory

  • Dataset with annotations and semantic images

• Download the following files to the models/ directory

  • Akin trained checkpoint

• Install dependencies

  pip install -r requirements.txt

• To train run

  python main.py --train

---

Authors

👤 Nishit Gajjar

• Github: @nishit
• LinkedIn: @nishit

👤 Vinoth Pandian

• Website: vinoth.info
• Github: @vinothpandian
• LinkedIn: @vinothpandian

---

Thanks to

• Leafinity: Code based on their implementation of SAGAN