Skip to content

[BMVC 2023 (Oral)] SketchDreamer: Interactive Text-Augmented Creative Sketch Ideation

License

Notifications You must be signed in to change notification settings

WinKawaks/SketchDreamer

Repository files navigation

SketchDreamer: Interactive Text-Augmented
Creative Sketch Ideation

PyTorch Conference Paper

Artificial Intelligence Generated Content (AIGC) has shown remarkable progress in generating realistic images. However, in this paper, we take a step "backward" and address AIGC for the most rudimentary visual modality of human sketches. Our objective is on the creative nature of sketches, and that creative sketching should take the form of an interactive process. We further enable text to drive the sketch ideation process, allowing creativity to be freely defined, while simultaneously tackling the challenge of "I can't sketch". We present a method to generate controlled sketches using a text-conditioned diffusion model trained on pixel representations of images. Our proposed approach, referred to as SketchDreamer, integrates a differentiable rasteriser of Bezier curves that optimises an initial input to distil abstract semantic knowledge from a pretrained diffusion model. We utilise Score Distillation Sampling to learn a sketch that aligns with a given caption, which importantly enable both text and sketch to interact with the ideation process. Our objective is to empower non-professional users to create sketches and, through a series of optimisation processes, transform a narrative into a storyboard by expanding the text prompt while making minor adjustments to the sketch input. Through this work, we hope to aspire the way we create visual content, democratise the creative process, and inspire further research in enhancing human creativity in AIGC.

overview

Citation

@inproceedings{qu2023sketchdreamer,
  title={SketchDreamer: Interactive Text-Augmented Creative Sketch Ideation},
  author={Qu, Zhiyu and Xiang, Tao and Song, Yi-Zhe},
  booktitle={BMVC},
  year={2023}
}

Instructions

Dependencies

This repository is based on CLIPasso and Stable-Dreamfusion. We would like to thank the authors of these work for publicly releasing their code. Please follow the dependencies of CLIPasso to install DiffVG. After running CLIPasso successfully, please install the following libraries.

pip install --upgrade diffusers accelerate transformers

Script

sh run.sh

Parameters

The optimisation process is susceptible to the initialsation of Bezier curves. We support the following methods to set the positions of strokes.

Bounding box: follows the format "x1, y1, x2, y2". All parameters should be in the range 0 to 1.

# a square area with width and height of (0.4 * 512)
sh run.sh --bbox "0.3,0.3,0.7,0.7"

Coordinates: follows the format "x1,y1,x2,y2,...,xN,yN". All parameters should be in the range 0 to 1.

# 4 curves with the starting points located in 4 corners
sh run.sh --num_strokes 4 --init_point "0,0,0,1,1,0,1,1"

When the value of num_aug_clip is set to 1, the data augmentation will be canceled. Data augmentation will make each iteration significantly slower, but will produce better results.

About

[BMVC 2023 (Oral)] SketchDreamer: Interactive Text-Augmented Creative Sketch Ideation

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 3

  •  
  •  
  •