AI Paper Atlas

Research map / Detection and segmentation / CNN features

Cost Aggregation with 4D Convolutional Swin Transformer for Few-Shot Segmentation

Sunghwan Hong, Seokju Cho, Jisu Nam, Stephen Lin, Seungryong Kim

ECCV 2022 · Detection and segmentation › CNN features · approach: Plain and high-resolution ViTs · 152 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 Swin Transformer for cost aggregation to the baseline (ablation)VAT baseline · FSS-1000 1-shotmIoU: 80.0 -> 85.4Improved
replace non-overlapping VEM with overlapping-convolution VCM (ablation)VAT · FSS-1000 1-shotmIoU: 85.4 -> 87.0Improved
add guided pyramidal processing (GPP) (ablation)VAT · FSS-1000 1-shotmIoU: 87.0 -> 87.3Improved
add affinity-aware transformer decoder (ATD) (ablation)VAT · FSS-1000 1-shotmBA: 60.7 -> 68.0Improved
use transformers in the decoder instead of convolutions only (ablation)ATD · FSS-1000 1-shotmIoU: 87.3 -> 90.3Improved
use 4D Conv. Swin transformer instead of center-pivot 4D convolutions as aggregator (ablation)VAT · FSS-1000 1-shotmIoU: 88.1 -> 90.3Improved
use Swin aggregator instead of Linear transformer (ablation)VAT · FSS-1000 1-shotmIoU: 87.7 -> 90.3Improved
use VCM instead of VEM with Swin Transformer (ablation)VAT · FSS-1000 1-shotmIoU: 89.9 -> 90.3Improved
use VCM instead of VEM with Linear Transformer (ablation)Linear Transformer aggregator · FSS-1000 1-shotmIoU: 87.0 -> 87.7Improved
swap ResNet101 backbone for Swin transformer backbone (ablation)VAT · FSS-1000 1-shotmIoU: 90.3 -> 89.8No clear effect
add PFE-Net style data augmentation (ablation)VAT · PASCAL-5i mean mIoUmIoU: 67.9 -> 67.3Hurt
add CATs data augmentation (ablation)VAT · PASCAL-5i mean mIoUmIoU: 67.9 -> 66.1Hurt

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