Holger Winnemoeller

Recognizing Image Style.

The style of an image plays a significant role in how it is viewed, but style has received little attention in computer vision research. We describe an approach to predicting style of images, and perform a thorough evaluation of different image features for these tasks. We find that features learned in a multi-layer network generally perform best -- even when trained with object class (not style) labels. Our large-scale learning methods results in the best published performance on an existing dataset of aesthetic ratings and photographic style annotations. We present two novel datasets: 80K Flickr photographs annotated with 20 curated style labels, and 85K paintings annotated with 25 style/genre labels. Our approach shows excellent classification performance on both datasets. We use the learned classifiers to extend traditional tag-based image search to consider stylistic constraints, and demonstrate cross-dataset understanding of style.

DiscoverySpace: Crowdsourced Suggestions Onboard Novices in Complex Software

Complex software, while powerful for experts, can overwhelm new users. Novices often do not know how to execute tasks, what they want to achieve, or even what is possible. We aim to address these problems by leveraging the large expert user base such programs tend to have. We present DiscoverySpace, a prototype extension panel for Adobe Photoshop that suggests macros for effects to apply to the user¿s photo, based on features of the photo. These macros are Photoshop actions that have been created and shared online by the user community, and can be applied in one click. Our work demonstrates how high-level macro suggestions can help users get started in a complex application. Preliminary feedback from a pilot study indicates that these suggestions may be most useful for users with exploratory goals, such as ¿make this photo more fun,¿ rather than users with very specific goals.