4.6 Experimental Setup 73 (a) Square. (b) Arrow. (c) Airplane. (d) ModelNet40. (e) COIL-100. Figure 4.4: Example images from each of the datasets used. Each image corresponds to an example data point for a combination of two factors, e.g. colour and orientation. The factors change horizontally and vertically. The boundaries for the Square, Arrow and Airplane dataset are periodic. For the ModelNet40 and COIL-100 dataset, the vertical direction represents different object instances and the horizontal direction represents the rotation of the corresponding object. Arrow The Arrow dataset consists of a set of images depicting a coloured arrow at a given orientation. The dataset is generated by applying cyclic shifts of its colour and in-plane rotations. The cyclic colour shifts were obtained by pre-selecting a fixed set of 64 colours from a circular hue axis. The in-plane rotations were obtained by rotating the arrow along an axis perpendicular to the picture plane over 64 predefined positions.
RkJQdWJsaXNoZXIy MjY0ODMw