Popular apps such as Instagram, Cameraþ, EyeEm, Hipstamatic, and Aviary have introduced the idea of photographic filters. These filters, which can be complex, typically adjust the intensity, hue, saturation, or contrast in the image either locally or globally. They can also apply color lookup tables or overlay one or more masking filters such as a vignetting mask (darkening edges and corners), thereby generating the Polaroid effect, for example.
Different filters are best applied to different types of images to obtain aesthetically pleasing
pictures. Well-known examples of photographic filters provided by the Instagram app, for example, are the Rise filter, which works best for close-up portraits; the Hudson filter, which is best applied to photos of building exteriors; and the Lo-Fi filter for shots of food. A Remembering and then later finding the best working filter may be frustrating. For users to quickly share a picture with friends on Facebook or Snapchat, ideally the best filter would be applied automatically.
The recent breakthroughs in visual concept detection now allow for the reliable matching of scenes to the most appropriate filter in real time.11 This means that when the user points the camera at a scene of architecture, for example, the Impala app can apply the appropriate photographic filter, enhancing the image’s aesthetical appeal (see Figure 4).