Image organization based on image content

A computer-implemented method of organizing images based on faces of people depicted in the images, the method, comprising:

• detecting, by a processing unit of a computing device, the faces of people depicted in the images;
• extracting, by the processing unit, content including the detected faces from the images;
• computing, by the processing unit, similarity confidence data by comparing the extracted content, the computed similarity confidence data between the extracted content indicating a likelihood that the detected faces represent a same person;
• and
• generating, by the processing unit, a multi-dimensional distribution of the extracted content, the distribution including groupings of the extracted content based, at least in part, upon the computed similarity data;
• displaying, on a display of the computing device, the multi-dimensional distribution as a layout of the faces of the people depicted in the images such that the faces in the layout are displayed in multiple clusters of similar faces based on the groupings of the extracted content and such that distances between the faces in the layout are based on the computed similarity confidence data;
• and
• receiving, by a user input device of the computing device, user input to tag the clusters with identifiers for the people depicted in the images, the identifiers of tagged clusters automatically propagating to the images associated with the groupings of the extracted content.

One or more computer storage media storing computer-executable instructions that, when executed by a processor of a computing device, cause the computing device to perform a method for organizing images based on faces of people depicted in the images, the method comprising:

• detecting the faces of people depicted in the images;
• extracting content including the detected faces from the images;
• computing similarity confidence data by comparing the extracted content, the computed similarity confidence data between the extracted content indicating a likelihood that the detected faces represent a same person;
• generating a visual distribution of the extracted content, the visual distribution including groupings of the extracted content based, at least in part, upon the computed similarity confidence data;
• displaying the visual distribution as a multi-dimensional layout of the faces such that the faces in the layout are displayed in multiple clusters of similar faces based on the groupings of the extracted content and such that distances between the faces in the layout are based on the computed similarity confidence data;
• and
• receiving user input to tag the clusters with identifiers for the people depicted in the images, the identifiers of tagged clusters automatically propagating to the images associated with the groupings of the extracted content.