Method and apparatus for performing visual effects rendering on an image are described. User specifications are processed for at least one visual effect for the image. Scene analysis information associated with the image may be accessed, wherein the scene analysis information is based on the user specifications and indicates at least one pixel of the image for coupling with the at least one visual effect. The at least one pixel is accentuated using depth map information, wherein dimensions for the at least one pixel are determined in response to corresponding depth values of the at least one pixel. An output image is generated having at least one visual effect.