I am going through the equations of a research paper that describes the restoration of old films and came across the following equation:
$$y_t=D(s_t\frown\bar{s_t})\in\mathbb{R}^{W \times H}.$$
The result is a single recovered image matrix where the D represents a decoder network who's input is a concatenation of two encoded hidden states. I am having difficulties understanding what is meant by the second part of the equation.
My first thought was that it is constricting the bounds of the result to fit within the height and width of the original image that is being recovered.