SABR: Stable Adaptive Bitrate via Behavior-Cloning Pretrain + RL Fine-tune
Description
Network-aware ABR algorithm combining behavior-cloning pretraining with reinforcement-learning fine-tuning; companion code (SABR) and dataset (ABRBench) released.
Resource details
- Category path
- Players & Clients›Mobile Players
- Classification
- Not yet classified
SABR: Stable Adaptive Bitrate via Behavior-Cloning Pretrain + RL Fine-tune is cataloged in Players & Clients and Mobile Players.
URL
Related resources
- GreenABR+: Generalized Energy-Aware ABR for Mobile Video StreamingACM TOMM paper on generalized reinforcement-learning-based battery/energy-aware adaptive bitrate algorithm for mobile video streaming.
- RL-based ABR Evaluation on Commercial 5G Networks for UHD DeliveryPensieve-variant reinforcement-learning ABR algorithm implemented and evaluated on commercial 5G SA/NR-DC networks for 4K/8K mobile video d…
- NextPlayerNative Android video player with H.265/VP9/AV1 decode, HDR10+/Dolby Vision support, and modern Compose UI.
- pAV0NEAndroid multimedia player built with Kotlin + ExoPlayer using GAV1 decoder for AV1, with DASH support.
- moko-media — Kotlin Multiplatform Media LibraryKMP library for media picking and video player controls across Android/iOS.
- webrtc-externalsWebExtension replicating the chrome://webrtc-internals workflow cross-browser (Firefox/Edge included).
- webrtc-internals-exporterChromium extension exporting WebRTC stats to Prometheus PushGateway with an included Grafana dashboard.
- Google Cast — Official GitHub OrganizationOfficial Google Cast samples hub containing sender and receiver reference implementations across multiple platforms.
- videowall (info-beamer)Master/client Python setup for showing distinct fullscreen video per screen, e.g. one per Raspberry Pi, for signage installations.
- SmartTubeNextAdvanced open-source video player for Android TV and set-top box devices.