Memorable Maps: A Framework for Re-Defining Places in Visual Place Recognition

This paper presents a cognition-inspired agnostic framework for building map Visual Place Recognition. draws inspiration from human-memorability, utilizes the traditional image entropy concept and computes static content in an image; thereby presenting tri-folded criteria to assess ‘memorability’ of visual place recognition. A dataset namely ‘ESSEX3IN1’ is created, composed highly confusing images indoor, outdoor natural scenes analysis. When used conjunction with state-of-the-art recognition methods, proposed provides significant performance boost these techniques, as evidenced by results on ESSEX3IN1 other public datasets.