Autocomplete Repetitive Stroking with Image Guidance

Abstract

Kong(a) user input (b) suggestion (c) accept (d) type visualization (e) result Figure 1: Example of our system workflow. A user stipples over a leaf region of a reference image (a) while our system predicts what she might draw next (b) (blue strokes: inferred exemplars; pale red region: inferred target region; semi-transparent strokes: system suggestions), which is then accepted by the user (c) (green strokes: user inputs or accepted suggestions in this scene). (d) visualizes all the manually drawn content in black (261 strokes) and autocompleted content in red (3510 strokes). (e) shows the final result with different repetitive stroke patterns over different regions. Our autocomplete system can reduce tedious repetitive inputs, while being fully under user control.