Reproduce “Low-Cost Deep Learning for Automated Keyframe Interpolation in Sprite Animation” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Kenneth Alonso, Gustavo Patow
DeepInterframes generates intermediate sprite frames between key poses using multiscale motion estimation, RGBA-aware processing, alpha-gated blending, and encoder-decoder skip connections to preserve detail. On a 42-triplet test set, it reports mean FLIP error of 0.258 ± 0.156, versus 0.313 ± 0.214 for RIFE. This lower mean does not isolate architecture from sprite-specific training data or establish whole-sequence consistency.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.