AI Paper Atlas

Research map / Label-efficient and robust learning / Knowledge distillation

Optical Flow Distillation: Towards Efficient and Stable Video Style Transfer

Xinghao Chen, Yiman Zhang, Yunhe Wang, Han Shu, Chunjing Xu, Chang Xu

ECCV 2020 · Label-efficient and robust learning › Knowledge distillation · approach: Few-shot and metric distillation · 62 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 residual distillation loss (ablation)Student from scratch · MPI Sintel 5 scenese_stab sum: 0.4019 -> 0.3424Improved
add low-rank distillation loss from input videos (ablation)Student from scratch · MPI Sintel 5 scenese_stab sum: 0.4019 -> 0.3855Improved
add low-rank loss to residual distillation loss (ablation)Student + L_res · MPI Sintel 5 scenese_stab sum: 0.3424 -> 0.3050Improved
distill low rank from input videos instead of teacher outputs (ablation)Student + L_res · MPI Sintel 5 scenese_stab sum: 0.3415 -> 0.3050Improved
add temporal loss to student baselineStudent from scratch · MPI Sintel 5 scenese_stab sum: 0.4019 -> 0.3787Improved
add residual distillation loss on top of temporal loss (ablation)Student w/ L_temp · MPI Sintel 5 scenese_stab sum: 0.3787 -> 0.3317Improved
add low-rank loss on top of temporal loss and residual loss (ablation)Student w/ L_temp + L_res · MPI Sintel 5 scenese_stab sum: 0.3317 -> 0.2938Improved
apply full optical flow distillation to student with temporal lossStudent w/ L_temp · MPI Sintel 5 scenese_stab sum: 0.3787 -> 0.2938Improved
remove optical flow from inference via distilled student instead of teacherTeacher · MPI Sintel 5 scenesFPS: 4.67 -> 183Improved
distill student instead of using teacher networkTeacher · MPI Sintel 5 scenese_stab sum: 0.3435 -> 0.2938Improved
increase K from 3 to 4 frames for low-rank loss (ablation)Student (Ours) · MPI Sintel 5 scenese_stab sum: 0.3158 -> 0.3061Improved
increase K from 4 to 5 frames for low-rank loss (ablation)Student (Ours) · MPI Sintel 5 scenese_stab sum: 0.3061 -> 0.3050No clear effect

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