Abstract
Techniques for adaptive real-time video communication are described. A device capability classifier, optionally based on a machine-learning model, determines codec eligibility beyond static device attributes. During a session, an endpoint measures encoding latency to adjust encoder complexity presets and to switch from a higher-complexity codec to a lower-complexity codec when real-time criteria are not satisfied. A receiving endpoint measures decoding latency and can signal the sender to change codec or configuration responsive to receiver-side performance. The techniques also support asymmetric operation in which an endpoint transmits using a first codec while receiving using a second codec. Additionally, temporal layering is enabled under network impairment (e.g., packet loss) and disabled upon recovery to improve compression efficiency when conditions are good.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Anonymous, "Adaptive Real-Time Video Communication with Latency-Aware Codec Selection and Asymmetric Codec Design", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/10610