Learning Condition-Invariant Scene Representations for Place Recognition across the Seasons Using Auto-Encoder and ICA

Figure 12

Obtained ICA filters (256) using the image patches of the summer-fall dataset.