During the acquisition and cross-media reproduction of videos, colors can be distorted because the color gamut of the camera is limited and may differ from the color gamut of the display device. Skin color is among the most sensitive colors to the human eye. Therefore, skin-color distortion can deteriorate viewers' video experience. Skin-color enhancement is a processing technology that adjusts a distorted skin color to improve the display quality. Particularly in video processing, the self-adaptability of a skin-color model must be improved while considering the real-time performance and computational load of the algorithms. For these purposes, the present paper proposes an adaptive skin-color enhancement method for real-time video processing. The update of the skin-color model is guided by shot boundaries, which can reduce the computational load of updating. Second, a dynamic skin-color model updated with the shot boundary is built for skin detection. Finally, the preferred skin-color model and skin-color enhancements for different races are achieved through subjective experiments. The proposed method achieved higher mean opinion scores than the existing methods in subjective evaluation experiments. In addition to achieving the targeted skin-color enhancement, the proposed method significantly reduced the computational load of model update.