Self-Supervised 3-D Semantic Representation Learning for Vision-and-Language Navigation | IEEE Journals & Magazine | IEEE Xplore