Hyperstroke: A Novel High-quality Stroke Representation for Assistive Artistic Drawing

TL;DR

Hyperstroke introduces a new stroke representation capturing detailed visual and temporal features, enabling improved assistive drawing tools through a transformer-based model that learns from real artistic videos.

Contribution

It presents hyperstroke, a novel stroke encoding method using vector quantization, and a transformer-based framework for assistive drawing that models intricate stroke details and temporal dynamics.

Findings

Effective modeling of fine stroke details including RGB and opacity.

Successful learning of stroke representations from real artistic videos.

Enhanced assistive drawing capabilities demonstrated in exploratory evaluation.

Abstract

Assistive drawing aims to facilitate the creative process by providing intelligent guidance to artists. Existing solutions often fail to effectively model intricate stroke details or adequately address the temporal aspects of drawing. We introduce hyperstroke, a novel stroke representation designed to capture precise fine stroke details, including RGB appearance and alpha-channel opacity. Using a Vector Quantization approach, hyperstroke learns compact tokenized representations of strokes from real-life drawing videos of artistic drawing. With hyperstroke, we propose to model assistive drawing via a transformer-based architecture, to enable intuitive and user-friendly drawing applications, which are experimented in our exploratory evaluation.