Camel: Frame-Level Bandwidth Estimation for Low-Latency Live Streaming under Video Bitrate Undershooting
This new algorithm is fixing the stuttering that plagues every major live stream.
Deep Dive
Researchers have unveiled 'Camel,' a new frame-level congestion control algorithm designed to eliminate stalling in low-latency live streams. It solves a key problem where traditional methods misjudge bandwidth, causing frequent buffering even when capacity is available. In a massive real-world test on a platform with 250M users, Camel increased 1080P streams by 70.8%, boosted media bitrate by 14.4%, and reduced stalling by up to 14.1%.
Why It Matters
This breakthrough could make choppy, buffering live streams on platforms like Twitch and YouTube a thing of the past for billions of viewers.