Scene Categorization with Phase-randomized Masking
Yes, you are CORRECT! It was a Forest!
Our results show that this type of mask produces an intermediate degree of scene gist masking. Phase-randomized image masks cause much stronger masking than white noise masks (Loschky et al., 2007; Loschky et al., 2010). This is because phase-randomized masks share identical spatial frequency and orientation information with scenes, whereas the white noise masks do not. This suggests that spatial frequency and orientation information masks scene gist, and is therefore useful for recognizing scene gist. Other research from lab has shown that the distribution of spatial frequencies is more important than that of their orientations for masking scene gist (Hansen & Loschky, 2013). Our results also show, however, that this usefulness is somewhat limited compared to masks containing local phase+amplitude information, such as edges and lines (Hansen & Loschky, 2013; Loschky et al., 2010).
