# Content Adaptive Encoding For Interactive Game Streaming

arXiv: https://arxiv.org/abs/2511.22327 · HTML: https://mlsanity.com/paper/2511.22327/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 11 of its 26 references that are in the map.
- First posted on arXiv 27 Nov 2025.

## Builds on (most cited first, 11 of 11)

- [Efficient Bitrate Ladder Construction for Content-Optimized Adaptive Video Streaming](https://mlsanity.com/paper/2102.04550/) (2021)
- [GAMIVAL: Video Quality Prediction on Mobile Cloud Gaming Content](https://mlsanity.com/paper/2305.02422/) (2023)
- [Deep Video Precoding](https://mlsanity.com/paper/1908.00812/) (2019)
- [User Generated HDR Gaming Video Streaming: Dataset, Codec Comparison and Challenges](https://mlsanity.com/paper/2103.02189/) (2021)
- [Convex Hull Prediction for Adaptive Video Streaming by Recurrent Learning](https://mlsanity.com/paper/2206.04877/) (2022)
- [Ensemble Learning for Efficient VVC Bitrate Ladder Prediction](https://mlsanity.com/paper/2207.10317/) (2022)
- [Study of Compression Statistics and Prediction of Rate-Distortion Curves for Video Texture](https://mlsanity.com/paper/2102.04167/) (2021)
- [Bitrate Ladder Construction using Visual Information Fidelity](https://mlsanity.com/paper/2312.07780/) (2023)
- [On the benefit of parameter-driven approaches for the modeling and the prediction of Satisfied User Ratio for compressed video](https://mlsanity.com/paper/2206.09854/) (2022)
- [Convex Hull Prediction Methods for Bitrate Ladder Construction: Design, Evaluation, and Comparison](https://mlsanity.com/paper/2310.15163/) (2023)
- [Convex-hull Estimation using XPSNR for Versatile Video Coding](https://mlsanity.com/paper/2406.13712/) (2024)

## Abstract (arXiv)

Video-on-demand streaming has benefitted from textitcontent-adaptive encoding (CAE), i.e., adaptation of resolution and/or quantization parameters for each scene based on convex hull optimization. However, CAE is very challenging to develop and deploy for interactive game streaming (IGS). Commercial IGS services impose ultra-low latency encoding with no lookahead or buffering, and have extremely tight compute constraints for any CAE algorithm execution. We propose the first CAE approach for resolution adaptation in IGS based on compact encoding metadata from past frames. Specifically, we train a convolutional neural network (CNN) to infer the best resolution from the options available for the upcoming scene based on a running window of aggregated coding block statistics from the current scene. By deploying the trained CNN within a practical IGS setup based on HEVC encoding, our proposal: (i) improves over the default fixed-resolution ladder of HEVC by 2.3 Bjontegaard Delta-VMAF points; (ii) infers using 1ms of a single CPU core per scene, thereby having no latency overhead.