Fig. 5
From: Conscious perception of natural images is constrained by category-related visual features
Target similarity between T1 and T2 explains T2 performance. a Representational sampling was used to construct trials of experiment 2. Each of the sixteen T2s were either preceded by a T1 from the same/different category and similar/dissimilar in representational space within layer 5 of the DCNN. To ensure that participants did not use low-level statistics (such as colour) when reporting the targets, we switched the response menu to a semantic task. b Behavioural results from experiment 2. Our results show that features similarity explain a significant portion of T2 performance. Individual dots correspond to individual subjects. Error bars indicate the 95% confidence interval. Statistics were performed using a repeated measures ANOVA (see results). *p < 0.05, **p < 0.01, ***p < 0.001
