Figure 2
From: Prioritization of emotional faces is not driven by emotional content
Results for decoding which image is detected first for faces with different expressions (top row) and for faces that have the same type of expression (bottom row). (A,C) The pool of decoding performances for the feature selection and the control models. On their y-axes is the fraction of models that show a particular performance. Chance performances from the control models are shown in purple, performances of the feature selection models are shown in red. (B,D) The locations of the features that were found to be significant for decoding within Fourier space for the images to the right and to the left of fixation separately. Note that in these Fourier space images, the angle between a point and the centre indicates the propagation direction of the waveform. As such, points on the horizontal axis indicate light–dark changes from left to right and therefore correspond to contrasts in vertical orientations within the images. The coloured areas in the Fourier spaces show the locations of the significant features with their colour reflecting their associated performances in percentage correct as shown in the colours on the right.
