Real-time Vision-Language-Navigation based on a Lite Pre-training Model | IEEE Conference Publication | IEEE Xplore