Style Transfer for Headshot Portraits

Teaser We transfer the styles from the example portraits in the insets in (b) to the input in (a). Our transfer technique is local and multi-scale, and tailored for headshot portraits. First, we establish a dense correspondence between the input and the example. Then, we match the local statistics in all different scales to create the output. Examples from left to right: image courtesy of Kelly Castro, Martin Schoeller, and Platon.


YiChang Shih, Sylvain Paris, Connelly Barnes, William T. Freeman, and Fr├ędo Durand, Style Transfer for Headshot Portraits, to appear in SIGGRAPH 2014

Paper: High resolution PDF (116 MB) | Low resolution PDF (12 MB)


Headshot portraits are a popular subject in photography but to achieve a compelling visual style requires advanced skills that a casual photographer will not have. Further, algorithms that automate or assist the stylization of generic photographs do not perform well on headshots due to the feature-specific, local retouching that a professional photographer typically applies to generate such portraits. We introduce a technique to transfer the style of an example headshot photo onto a new one. This can allow one to easily reproduce the look of renowned artists. At the core of our approach is a new multiscale technique to robustly transfer the local statistics of an example portrait onto a new one. This technique matches properties such as the local contrast and the overall lighting direction while being tolerant to the unavoidable differences between the faces of two different people. Additionally, because artists sometimes produce entire headshot collections in a common style, we show how to automatically find a good example to use as a reference for a given portrait, enabling style transfer without the user having to search for a suitable example for each input. We demonstrate our approach on data taken in a controlled environment as well as on a large set of photos downloaded from the Internet. We show that we can successfully handle styles by a variety of different artists.

Video (no audio, best viewed at 1920x1080)

Additional results

See our method on 94 images in Flickr dataset at here.

Data & code

  • Code (25MB, MatLab, containing pre-built binary test on Debian)
  • Data (679MB, containing examples, Flickr dataset, and inputs in our paper)
  • Supplemental materials

  • More comparisons at the supplemental document.
  • Press


    We thank photographers Kelly Castro, Martin Schoeller, and Platon for allowing us to use their photographs in the paper. We also thank Kelly Castro for discussing with us how he works and for his feedback, Michael Gharbi and Krzysztof Templin for being our portrait models. We acknowledge the funding from Quanta Computer and Adobe.