Over the last decade, research in Human Computer Interaction has focused on the development of interfaces that leverage the users' pre-existing skills and expectations from the real world, rather than requiring them to adapt to the constraints of technology driven design. In the context of remote collaboration or communication interfaces, the ultimate goal has been to develop interfaces that will allow remote participants to interact with each other in a human sense, as if they were co-located or in a face-to-face meeting. Research in social psychology has shown that the face is an important channel in non-verbal communication and real world interactions. Non-verbal cues that come from the face are the basis for building trust and professional intimacy and are critical for collaboration, negotiation, persuasion and communication. This research investigated the challenges of bringing non-verbal cues conveyed by the face into a communication interface. To meet these challenges, the proposed system allowed participants to convey the most distinctive nonverbal cues by using three different modes; point cloud, dynamic texture mapping and geometric deformation. A human factor evaluation was undertaken to find out how realistically these non-verbal cues could be expressed by the personalized avatar of the participant.