GamingVideoSET
A 10-bit UHD-HDR gaming video dataset for video quality assessment, QoE prediction, and encoding optimization research.
Overview
GamingHDRVideoSET is a specialized dataset designed for research on gaming video streaming applications. It addresses the unique challenges of live gaming video encoding, which differs from traditional VoD services due to real-time constraints and artificial scene content. The dataset enables comparable and generalizable research in video quality assessment and Quality of Experience (QoE) modeling for the gaming streaming domain.
Key Features
- High-Quality Source Videos — 18 reference gaming videos at 3840x2160 resolution, 10-bit HDR, and 30 fps.
- Multiple Codec Distortions — 384 distorted sequences encoded with H.264, H.265, VP9, and AV1 codecs at four bitrate levels.
- Objective Quality Metrics — Includes PSNR and HDR-VQM scores for all encoded video sequences.
- High Frame Rate Content — Additional six gaming videos at 60 fps from three different games.
- Research-Focused Design — Created specifically for video quality assessment, QoE prediction, and encoding optimization studies.
Philosophy
The dataset is built to provide an openly available, standardized resource that allows researchers to achieve comparable and generalizable results in gaming video streaming research, filling a gap for high-quality HDR gaming video data.
Related Projects
The Quick, Draw! DatasetDocumentation on how to access and use the Quick, Draw! Dataset.
StarDataStarcraft AI Research Dataset
GTA-3D DatasetA dataset of 2D imagery, 3D point cloud data, and 3D vehicle bounding box labels all generated using the Grand Theft Auto 5 game engine.
Pokémon GO Pokédex🐲 Pokédex of Pokémon GO in JSON.
Found a gem we're missing?
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.