Mechanism to validate and compare interdevice/ destination media synchronization solutions adopted in multi-screen scenarios
Keywords:computer vision, multimedia synchronization, OCR, image recognition
Simultaneous consumption of related content in one (or multiple) devices (i.e., multi-screen scenarios) is commonplace. To provide a satisfactory Quality of Experience (QoE) in this type of scenarios it is necessary to adopt synchronization mechanisms to keep the involved content playout processes in a synchronized state. Most of those mechanisms are software-based and usually perform some estimations, at runtime, of the playout latency. Therefore, the accuracy of the measured synchronization level can be difficult to evaluate or assess. In this work, a computer vision-based mechanism to evaluate the achieved accuracy in multi-screen scenarios is proposed. Thus, the accuracy level which is obtained through the synchronization mechanisms implemented in those systems or applications can be assessed at the end point of the content presentation. Additionally, in order to show the utility of the proposed mechanism, some evaluations have been conducted for a common use case, a videowall, in which accurate synchronization is a requirement.