Bitrate-Ladder-Benchmark — learning-based ladder prediction

Companion code benchmarking learning-based bitrate ladder prediction methods for adaptive streaming (Telili et al.).

Link

• Visit resource

Related resources

• ABR-ArenaETH NSG global testing infrastructure to evaluate and compare adaptive bitrate algorithms across real-world environments, based on Puffer.
• SABR — Stable ABR Framework (BC + RL)Stable adaptive bitrate framework combining behavior-cloning pretraining with RL fine-tuning (arXiv 2509.10486).
• per-title-analysis — Per-Title Bitrate Ladder OptimizerPython per-title encoding ladder optimizer using CRF and metric (SSIM/PSNR) analyzers, Netflix-style, on ffmpeg/ffprobe.
• encodex — AI content-aware convex-hull encoderAI-agent content-aware encoder generating convex-hull Pareto-frontier bitrate ladders with content-complexity adjustments.
• ABRBench — ABR Benchmark with OOD Trace SetsBenchmark suite for ABR algorithms with 3G/4G+ network trace sets and out-of-distribution test splits.
• Cloudflare: MoQ — refactoring real-time media stackTechnical writeup on Media over QUIC and Cloudflare's first production MoQ relay network for real-time media.
• cloudflare/moq-rs — Media over QUIC in RustCloudflare's production-tested Rust implementation of IETF Media over QUIC Transport (draft-14), providing moq-relay, moq-api and moq-pub c…
• Meetecho: Bridging MoQ and WebRTCTechnical blog detailing a Janus plugin bridging WHIP/WHEP to Media over QUIC, with architecture details.
• rutgersc/m3u8-rs — Rust HLS Manifest ParserActive nom-based Rust parser for HLS m3u8 playlists (master and media).
• simple-whip-server (Janus)Node.js WHIP server implementation built on the Janus WebRTC gateway, reference implementation by Meetecho (Janus maintainers).