Fig. 3: Marker-to-marker translation in simulation.
From: Training tactile sensors to learn force sensing from each other
A Pipeline for marker deformation simulation. Quantitative evaluation of marker-to-marker translation via heatmaps of (B) FID and (C) KID, comparing performance before (i) and after (ii) applying the M2M model across all 132 sensor combinations. D t-SNE feature space visualization illustrating the alignment between target images and generated images across all sensor combinations, before (i) and after (ii) applying the M2M model.
