VANP: Learning Where to See for Navigation with Self-Supervised Vision-Action Pre-Training

Open in new window