Multimodal Human-Robot Collaboration: Combining Gaze And Pose Towards Fluent And Low Workload Interactions