The goal of this research is to design a new method of annotating traditional Chinese paintings, particularly by focusing on two issues that are unique to the Chinese masterpieces: (1) their form (“long scrolls versus canvases of regular size”) and (2) the “composition techniques for spatial recession (moving focus versus perspective).” A main unique technical aspect of the authors’ work is “how [they] infer the positions of the annotations in the 3D perception space of the image” since audio annotation of traditional Chinese paintings by itself is not new.

Unlike previous research on 3D digital exhibition techniques, the authors of this paper provide a way “to render 3D scenes and animations interpreted by modern designers, rather than [simply showing] the original art pieces.” As the user navigates the gigapixel image, a “user interface estimates the 3D position of the user’s current viewpoint, and ... synthesizes a stereo-audio field according to the viewer’s spatial orientation and distance from the annotations.” An algorithm is proposed to “obtain the 3D positions of all the sound sources.”

However, there are some issues. First, in the last paragraph of section 5.2, the authors indicate that the story “will not stop until it reaches the end.” I am curious to know whether it also has to do with the visitors’ change of viewpoints. Second, in section 6, the parameter K is defined as an adjustable constant. How does the range of K’s value affect the system and user experience? Finally, the evaluation of the proposed method is not very clear. For instance, how does it enhance visitors’ experience with the Chinese painting? Is there any other metric that can be adopted for evaluation, for example, the sound quality or the real-time rendering quality?