MultiRes-netVLAD: Augmenting place recognition training with low-resolution imagery

Publication
IEEE Robotics and Automation Letters