Supplementary Data for: AI Frame Generation on Budget Hardware: An Empirical Analysis of Performance, and Latency
收藏DataCite Commons2026-05-03 更新2026-05-07 收录
下载链接：
https://zenodo.org/doi/10.5281/zenodo.20014747
下载链接
链接失效反馈官方服务：
资源简介：
This dataset contains the supplementary benchmarking data, custom Python scripts, and video comparisons for the research paper "AI Frame Generation on Budget Hardware: An Empirical Analysis of Performance, and Latency." It provides a comprehensive evaluation of Real-Time Intermediate Flow Estimation (RIFE) models applied to budget-tier computing environments, specifically comparing dedicated (NVIDIA RTX 3050A) and integrated (Intel UHD Graphics) GPU performance.
The uploaded archive includes:
Custom Software: A Python-based Blender addon developed to interface the RIFE AI model directly with the rendering pipeline.
Raw Telemetry: Performance statistics tracking CPU/GPU utilization, framerates (FPS), system latency, and thermal metrics across Live and VFX rendering workloads.
Visual Benchmarks: High-resolution 2x (60 FPS) and extreme 6x (180 FPS) scalar video interpolations generated from a 30 FPS baseline for visual artifact and latency analysis.
Project Files: Source files and isolated hardware testing environments used to generate the control footage.
提供机构：
Zenodo
创建时间：
2026-05-03
