Between the Frames - Evaluation of Various Motion Interpolation Algorithms to Improve 360° Video Quality
This dataset evaluates three motion interpolation algorithms (ffmpeg MCI, Butterflow, and Super-SloMo) applied to 360° omnidirectional video content. The research processed 15 videos of 20 seconds duration, upscaling from 30 fps to 90 fps for HTC Vive Pro compatibility.
Description
This dataset evaluates three motion interpolation algorithms (ffmpeg MCI, Butterflow, and Super-SloMo) applied to 360° omnidirectional video content. The research processed 15 videos of 20 seconds duration, upscaling from 30 fps to 90 fps for HTC Vive Pro compatibility.
The dataset includes results from two subjective quality assessment tests with 24 and 27 participants respectively, using Modified Paired Comparison and Absolute Category Rating methodologies across 45-60 stimuli each. The research demonstrates that interpolated versions achieved significantly elevated quality scores, with Super-SloMo outperforming alternative algorithms when adequate computational resources are available.
The complete dataset (800.9 MB) contains the processed 360° video content and subjective test results. This work was conducted by the Audiovisual Technology Group at TU Ilmenau, a leading research center for 360° video and immersive media quality assessment, with contributions to ITU-T standards for subjective evaluation of 360° videos.
Access
Openly available for download from Zenodo
License
Creative Commons Attribution 4.0 International (CC-BY-4.0)