Adaptive Video Encoding for Different Video Codecs

By 2022, we expect video traffic to reach 82% of the total internet traffic. Undoubtedly, abundance video-driven applications will likely lead percentage a further increase in near future, enabled by associate advances devices' capabilities. In response this ever-growing demand, Alliance for Open Media (AOM) and Joint Video Experts Team (JVET) have demonstrated strong renewed interest developing new codecs. fast-changing codecs' landscape, there is thus, genuine need develop adaptive methods that can be universally applied different study, formulate encoding as multi-objective optimization process where quality (as function VMAF PSNR), bitrate demands, rate (in encoded frames per second) are jointly optimized, going beyond standard approaches focus on control targeting specific bandwidths. More specifically, create dense space (offline) then employ regression generate forward prediction models each one afore-described objectives, using only Pareto-optimal points. We demonstrate our approach leverages generated qualify real-time adaptation codecs (e.g., SVT-AV1 x265) variety datasets resolutions. To motivate establish promise future fast VVC encoders, also perform comparative performance evaluation both subjective objective metrics report savings among all possible pairs between VVC, SVT-AV1, x265, VP9