• Corpus ID: 245131058

Exploring Latent Dimensions of Crowd-sourced Creativity

@article{Kocasari2021ExploringLD,
  title={Exploring Latent Dimensions of Crowd-sourced Creativity},
  author={Umut Kocasari and Alperen Bag and Efehan Atici and Pinar Yanardag},
  journal={ArXiv},
  year={2021},
  volume={abs/2112.06978}
}
Recently, the discovery of interpretable directions in the latent spaces of pre-trained GANs has become a popular topic. While existing works mostly consider directions for semantic image manipulations, we focus on an abstract property: creativity. Can we manipulate an image to be more or less creative? We build our work on the largest AI-based creativity platform, Artbreeder, where users can generate images using pre-trained GAN models. We explore the latent dimensions of images generated on… 

Figures from this paper

One Artist's Personal Reflections on Methods and Ethics of Creating Mixed Media Artificial Intelligence Art

I intend to make a scientific contribution of my subjective ex- perience as a single unit of self-described “artist” leveraging artificial intelligence as an assistive visual creation tool, in the

References

SHOWING 1-10 OF 11 REFERENCES

GANalyze: Toward Visual Definitions of Cognitive Image Properties

A framework that uses Generative Adversarial Networks (GANs) to study cognitive properties like memorability is introduced and it is demonstrated that the same framework can be used to analyze image aesthetics and emotional valence.

The power of pictures: using ML assisted image generation to engage the crowd in complex socioscientific problems

This work weaving in carefully structured design elements to transform the activity of ML-assisted imaged generation into a catalyst for large-scale popular dialogue on complex socioscientific problems such as the United Nations Sustainable Development Goals (SDGs) and as a gateway for public participation in research.

Large Scale GAN Training for High Fidelity Natural Image Synthesis

It is found that applying orthogonal regularization to the generator renders it amenable to a simple "truncation trick," allowing fine control over the trade-off between sample fidelity and variety by reducing the variance of the Generator's input.

Graph2Pix: A Graph-Based Image to Image Translation Framework

A novel graph-to-image translation model called Graph2Pix, which takes a graph and corresponding images as input and generates a single image as output, is proposed, which is able to outperform several image- to- image translation frameworks on benchmark metrics.

DeepFakes: a New Threat to Face Recognition? Assessment and Detection

This paper presents the first publicly available set of Deepfake videos generated from videos of VidTIMIT database, and demonstrates that GAN-generated Deep fake videos are challenging for both face recognition systems and existing detection methods.

Visualizing Data using t-SNE

A new technique called t-SNE that visualizes high-dimensional data by giving each datapoint a location in a two or three-dimensional map, a variation of Stochastic Neighbor Embedding that is much easier to optimize, and produces significantly better visualizations by reducing the tendency to crowd points together in the center of the map.

Toward Quantifying Ambiguities in Artistic Images

It is shown that text processing of viewer responses can provide a fine-grained way to measure and describe image ambiguities.

Mask R-CNN

This work presents a conceptually simple, flexible, and general framework for object instance segmentation that outperforms all existing, single-model entries on every task, including the COCO 2016 challenge winners.

Measuring colorfulness in natural images. In Human vision and electronic imaging VIII, volume 5007, pages 87–95

  • International Society for Optics and Photonics,
  • 2003