<![CDATA[This study evaluates the performance of the Real-Time Intermediate Flow Estimation (RIFE) algorithm in enhancing the temporal continuity and perceptual quality of 3D-rendered animation. Using 13 animation samples from the short film “I Draw It”, the algorithm was tested at 480p, 720p, and 1080p resolutions to interpolate motion from 24 fps to 48 fps, increasing temporal density while preserving visual coherence. The assessment integrates objective measures Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM), and processing time with expert-based evaluations using a Likert-scale instrument to capture perceptual fidelity and motion smoothness. Experimental results indicate that higher resolutions consistently provide superior interpolation stability and detail preservation, with 1080p yielding the highest perceptual performance despite increased computational requirements. Conversely, 720p demonstrates a practical trade-off between visual quality and processing efficiency, making it suitable for preview stages in production workflows. This research addresses a gap in the literature by conducting a resolution-specific evaluation of RIFE within fully 3D-rendered animation, a context that remains underexamined in existing studies. The findings establish RIFE as a viable and reliable interpolation method for modern animation pipelines and highlight its potential for integration into real-time frameworks such as Unreal Engine and Unity, thereby supporting improved production efficiency and visually consistent output.]]>
Copyrights © 2026
