We

In this paper, we tackle the problem of associating combinations of colors to abstract categories (e.g. capricious, classic, cool, delicate, etc.). It is evident that such concepts would be difficult to distinguish using single colors, therefore we consider combinations of colors or color palettes. We leverage two novel databases for color palettes and we learn categorization models using low and high level descriptors. Preliminary results show that Fisher representation based on GMMs is the most rewarding strategy in terms of classification performance over a baseline model. We also suggest a process for cleaning weakly annotated data, whilst preserving the visual coherence of categories. Finally, we demonstrate how learning abstract categories on color palettes can be used in the application of color transfer, personalization and image re-ranking.

No abstract

No abstract

This paper introduces a novel approach to automatic concept transfer; examples of concepts are "romantic", "earthy", and "luscious". The approach modifies the color content of an input image given only a concept specified by a user in natural language, thereby requiring minimal user input. This approach is particularly useful for users who are aware of the message they wish to convey in the transferred image while being unsure of the color combination needed to achieve the corresponding transfer. The user may adjust the intensity level of the concept transfer to his/her liking with a single parameter. The proposed approach uses a convex clustering algorithm, with a novel pruning mechanism, to automatically set the complexity of models of chromatic content. It also uses the EarthMover's Distance to compute a mapping between the models of the input image and the target chromatic concept. Results show that our approach yields transferred images which effectively represent concepts, as confirmed by a user study.