AI Paper Atlas

Research map / Video / Video segmentation and tracking

Video Restoration Framework and Its Meta-Adaptations to Data-Poor Conditions

Prashant W Patil, Sunil Gupta, Santu Rana, Svetha Venkatesh

ECCV 2022 · Video › Video segmentation and tracking · approach: Video restoration with temporal priors · 10 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 scale modulation to the baseline network (ablation)Proposed video restoration network · REVIDE video de-hazingPSNR: 22.74 -> 23.69Improved
add temporal modulation to the baseline network (ablation)Proposed video restoration network · REVIDE video de-hazingPSNR: 22.74 -> 23.95Improved
combine scale and temporal modulation (STFM) (ablation)Proposed video restoration network · REVIDE video de-hazingPSNR: 22.74 -> 24.64Improved
add multi-level feature aggregation (MFA) (ablation)Proposed video restoration network · REVIDE video de-hazingPSNR: 24.64 -> 25.42Improved
add previous frame output feedback to the decoder (ablation)Proposed video restoration network · REVIDE video de-hazingPSNR: 25.42 -> 26.36Improved
add previous frame output feedback to the decoder (ablation)Proposed video restoration network · REVIDE video de-hazingSSIM: 0.8814 -> 0.9044Improved
use meta-learning adaptation instead of scratch trainingProposed network · night-time de-hazingPSNR: 21.99 -> 23.67Improved
use meta-learning adaptation instead of scratch trainingProposed network · night-time de-rainingPSNR: 23.57 -> 24.71Improved
use meta-learning adaptation instead of scratch trainingProposed network · night-time de-raining with veilPSNR: 22.21 -> 23.49Improved
fine-tune day-time model on night data instead of scratch trainingProposed network · night-time de-rainingPSNR: 23.57 -> 22.78Hurt
fine-tune day-time model on night data instead of scratch trainingProposed network · night-time de-hazingPSNR: 21.99 -> 22.01No clear effect
use meta-adaptation with only 10% of night training samplesProposed network · night-time restoration tasksPSNR/SSIMImproved

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