AI Paper Atlas

Research map / Image generation / Editing and style transfer

ForkGAN: Seeing into the Rainy Night

Ziqiang Zheng, Yang Wu, Xinran Han, Jianbo Shi

ECCV 2020 · Image generation › Editing and style transfer · approach: Image-to-image translation · 111 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 fork-shape generator (one encoder, two decoders) (ablation)ForkGAN · AlderleyFID / SIFT: 146 / 4.26 -> 131 / 7.12Improved
add dilated residual blocks (ablation)Fork-shape generator · AlderleyFID / SIFT: 131 / 7.12 -> 113 / 8.12Improved
add multi-scale discriminator (ablation)Fork-shape+DRB · AlderleyFID / SIFT: 113 / 8.12 -> 95.3 / 9.14Improved
add perceptual loss Lper and domain classification loss Lcls (ablation)Fork-shape+DRB+MSD · AlderleyFID / SIFT: 95.3 / 9.14 -> 70.7 / 11.5Improved
add refinement stage at test time (ablation)ForkGAN · AlderleyFID / SIFT: 70.7 / 11.5 -> 61.2 / 12.1Improved
use separate translation and reconstruction decoders instead of shared decoders (ablation)ForkGAN · AlderleyFID / SIFT: 73.8 / 9.29 -> 61.2 / 12.1Improved
night-to-day translation with ForkGAN before segmentationDeepLabv3 · BDD100K night (137 images)mIoU: 7.03 -> 14.4Improved
night-to-day translation with ToDayGAN before segmentationDeepLabv3 · BDD100K night (137 images)mIoU: 7.03 -> 8.19Improved
night-to-day translation with MUNIT before segmentationDeepLabv3 · BDD100K night (137 images)mIoU: 7.03 -> 2.44Hurt
augment day-only training with ForkGAN night translationsFaster R-CNN · BDD100K night detectionmAP: 22.1 -> 22.9Improved
augment day-only training with ToDayGAN translationsFaster R-CNN · BDD100K night detectionmAP: 22.1 -> 19.5Hurt
add ForkGAN translated night images to real night training dataFaster R-CNN · BDD100K night detectionmAP: 23.9 -> 26.2Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 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