Leveraging Multimodal Human Cues To Enhance Robot Learning From Demonstration