Real-Time video Communication (RTC) has been critical for teleconference, online education and many other emerging use cases. RTC is prone to fluctuations of the network conditions. The video compression bit rate, resolution and/or frame rate have to be reduced when severe packet loss and/or traffic congestion happens, resulting in degraded viewing experience. This talk will introduce how deep learning algorithms are designed and accelerated to counter these degradations in RTC.
Want to sponsor this event? contact us.