The present disclosure relates to a video conferencing system which provides non-obtrusive feedback to a presenter during a video conference to improve interactivity among participants of video conference. An image sensor, such as a camera, captures a sequence of images (i.e., a video) of the participant while a participant of the video conference is moving their port or a part of their body and sends the images to video conferencing server. The video conferencing server processes the images to recognize a type of gesture performed by the participant and selects an ambient graphic that corresponds to the recognized gesture. The video conferencing server sends the ambient graphic to a client device associated with the presenter. The client device associated with the presenter renders or displays the ambient graphic on a display screen of the client device without obscuring information displayed on the display screen of the client device.