AI Paper Atlas

Research map / Detection and segmentation / CNN features

Beyond Fixed Grid: Learning Geometric Image Representation with a Deformable Grid

Jun Gao, Zian Wang, Jinchen Xuan, Sanja Fidler

ECCV 2020 · Detection and segmentation › CNN features · approach: Context modeling for semantic segmentation · 26 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
replace fixed max pooling with grid max pooling on the deformed gridModified ResNet50 · Cityscapes semantic segmentation (1/8 downsampling)mIoU: 59.83 -> 64.75Improved
replace fixed average pooling with grid average poolingModified ResNet50 · Cityscapes semantic segmentation (1/32 downsampling)mIoU: 43.54 -> 48.30Improved
replace fixed pooling with grid pooling, measured on boundary qualityModified ResNet50 · Cityscapes semantic segmentation (1/16 downsampling)boundary F (4px): 49.00 -> 53.98Improved
add learned grid deformation instead of keeping the grid fixed (ablation)DefGrid annotation model · Cityscapes-MulticompF2: 74.77 -> 77.04Improved
add learned grid deformation instead of keeping the grid fixed (ablation)DefGrid annotation model · Cityscapes-MulticompmIoU: 81.67 -> 82.17Improved
reason on the deformed grid instead of Curve-GCN polygonCurve-GCN · ssTEM (cross-domain)F2: 59.85 -> 65.14Improved
reason on the deformed grid instead of Curve-GCN polygon, interactive annotationCurve-GCN · Cityscapes interactive annotation (round 0)F1: 58.01 -> 61.43Improved
predict per-cell labels on the deformed grid instead of pixel-wise labelspixel-wise annotation model · Cityscapes-StretchmIoU: 80.01 -> 81.42Improved
use L1 instead of L2 distance in reconstruction loss (ablation)DefGrid · training lossqualityImproved
add DefGrid for unsupervised superpixels versus SEAL supervisedsuperpixel baselines · BSDS500 testboundary precisionMixed

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 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