Enhancing content appearance

1. A method comprising:

receiving an image that is a first frame in a video stream, wherein the first frame is located at any point in the video stream;

selecting a reference object in the first frame, wherein the reference object is a first face, and wherein the selecting comprises:

determining a foreground and a background of the image by comparison of the first frame to one or more other frames in the video stream to identify the reference object based on a relative position of the reference object in the first frame and in each of the one or more other frames; and

selecting the reference object from the foreground;

determining one or more image parameter adjustments to apply to the first face to modify a skin tone of the first face, wherein the one or more image parameter adjustments meet predetermined face criteria, wherein the predetermined face criteria is based on one or more other faces in the one or more other frames, and wherein the one or more image parameter adjustments comprise adjustments to one or more of gamma, contrast, and brightness; and

applying the one or more image parameter adjustments to the first face in the first frame.

2. The method of claim 1 , wherein:

the predetermined face criteria is based on other images of the first face; and

selecting the reference object in the first frame further comprises analyzing at least one of a size of the reference object and a shape of facial features of the reference object.

3. The method of claim 1 , wherein the one or more image parameter adjustments include adjustments to one or more white balance parameters.

4. The method of claim 1 , further comprising applying the one or more image parameter adjustments to the one or more other frames in the video stream.

5. A method comprising:

receiving a first image that is a first frame in a video stream, wherein the first frame is located at any point in the video stream;

selecting a first face in the first frame by:

generating alpha values for the first frame;

applying foreground and background segmentation based on the alpha values to identify a foreground and a background; and

selecting the first face from the foreground;

determining one or more image parameter adjustments to apply to the first face to modify a skin tone of the first face based on a comparison of the first face to a reference image that includes a type of object corresponding to the first image, wherein the one or more image parameter adjustments comprise adjustments to one or more of gamma, contrast, brightness, luminosity, chrominance, red color, blue color, and green color; and

applying the one or more image parameter adjustments to the first face in the first frame.

6. The method of claim 5 , wherein:

generating the alpha values for the first frame includes assigning a first alpha value of the alpha values to a first object in the first frame;

applying foreground and background segmentation based on the alpha values comprises determining that the first alpha value exceeds an alpha value threshold and that the first object is part of the foreground; and

selecting the first face from the foreground includes determining that the first object is the first face.

7. The method of claim 5 , wherein the comparison of the first face to the reference image includes a determination of a relative position of one or more facial features in the reference image to corresponding one or more facial features in the first face.

8. The method of claim 5 , wherein the reference image is generated by compressing face data from a plurality of reference images into a composite representation and wherein the comparison of the first face to the reference image includes comparing the first face to the composite representation.

9. The method of claim 5 , wherein the one or more image parameter adjustments meet predetermined face criteria, and wherein the predetermined face criteria is based on other images of other faces.

10. The method of claim 5 , further comprising generating a likelihood that the first face in the first frame matches the reference image.

11. The method of claim 10 , wherein generating the likelihood is based on performing facial recognition on the first frame.

12. The method of claim 5 , wherein the one or more image parameter adjustments include adjustments to one or more white balance parameters.

13. The method of claim 5 , further comprising applying the one or more image parameter adjustments to one or more other frames in the video stream that are similar to the first frame.

14. A system comprising:

one or more processors; and

logic encoded in one or more tangible media for execution by the one or more processors and when executed operable to perform operations comprising:

receiving a first image that is a first frame in a video stream, wherein the first frame is located at any point in the video stream;

selecting a first face in the first frame by:

generating alpha values for the first frame;

applying foreground and background segmentation based on the alpha values to identify a foreground and a background; and

selecting the first face from the foreground;

determining one or more image parameter adjustments to apply to the first face to modify a skin tone of the first face based on a comparison of the first face to a reference image that includes a type of object corresponding to the first image, wherein the one or more image parameter adjustments comprise adjustments to one or more of gamma, contrast, brightness, luminosity, chrominance, red color, blue color, and green color; and

applying the one or more image parameter adjustments to the first face in the first frame.

15. The system of claim 14 , wherein:

generating the alpha values for the first frame includes assigning a first alpha value of the alpha values to a first object in the first frame;

applying the foreground and background segmentation based on the alpha values comprises determining that the first alpha value exceeds an alpha value threshold and the first object is part of the foreground; and

selecting the first face from the foreground includes determining that the first object is the first face.

16. The system of claim 14 , wherein the comparison of the first face to the reference image includes a determination of a relative position of one or more facial features in the reference image to corresponding one or more facial features in the first face.

17. The system of claim 14 , wherein the reference image is generated by compressing face data from a plurality of reference images into a composite representation and wherein the comparison of the first face the reference image includes comparing the first face to the composite representation.

18. The system of claim 14 , wherein the logic when executed is further operable to perform operations comprising generating a likelihood that the first face in the first frame matches the reference image.

19. The system of claim 14 , wherein the one or more image parameter adjustments include adjustments to one or more white balance parameters.

20. The system of claim 14 , wherein the logic when executed is further operable to perform operations comprising applying the one or more image parameter adjustments to one or more other frames in the video stream that are similar to the first frame.