AI Paper Atlas

Research map / Video / Video segmentation and tracking

FILM: Frame Interpolation for Large Motion

Fitsum Reda, Janne Kontkanen, Eric Tabellion, Deqing Sun, Caroline Pantofaru, Brian Curless

ECCV 2022 · Video › Video segmentation and tracking · approach: Video frame interpolation · 199 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add weight sharing across scales (ablation)FILM-med · Vimeo-90KPSNR: 35.28 -> 35.30Improved
add weight sharing across scales (ablation)FILM-lite · Vimeo-90KPSNR: 34.93 -> 35.09Improved
add weight sharing across scales (ablation)FILM (full) · Vimeo-90Ktraining stability: N/A -> 36.06Improved
train with style loss L_S instead of L1FILM · Xiph-4KPSNR: 33.78 -> 33.29Hurt
train with style loss L_S instead of L1FILM · Vimeo-90KPSNR: 36.06 -> 35.87Hurt
add VGG perceptual loss to L1FILM · Vimeo-90KPSNR: 36.06 -> 35.76Hurt
add Gram matrix style loss on top of VGG lossFILM · Vimeo-90KPSNR: 35.76 -> 35.87Improved
train on custom large-motion dataset instead of Vimeo-90KFILM · Xiph-4KPSNR: +0.5dBImproved
increase training motion range (0-40 up to 0-120 px) (ablation)FILM-med · Vimeo-90KPSNR: PSNR goes down significantlyHurt
train on motion range 0-100 px (ablation)FILM-med · Xiph-4KPSNR: 0-100 pixelsImproved
use FILM-L_S rather than SoftSplat-L_F (perceptual losses)FILM · Xiph-4KPSNR: 32.55 -> 33.29Improved
use FILM instead of ABME for inference speedFILM · 720p inferenceinference time (s): 1.554 -> 0.393Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy