Eye contact detection via deep neural networks

With the presence of ubiquitous devices in our daily lives, effectively capturing and managing user attention becomes a critical device requirement. While gaze-tracking is typically employed to determine the user’s focus of attention, gaze-lock detection to sense eye-contact with a device is proposed in[16]. This work proposes eye contact detection using deep neural networks, and makes the following contributions: (1) With a convolutional neural network (CNN) architecture, we achieve superior eye-contact detection performance as compared to [16] with minimal data pre-processing; our algorithm is furthermore validated on multiple datasets, (2) Gaze-lock detection is improved by combining head pose and eye-gaze information consistent with social attention literature, and (3) We demonstrate gaze-locking on an Android mobile platform via CNN model compression.