AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

FASA: Feature Augmentation and Sampling Adaptation for Long-Tailed Instance Segmentation

Yuhang Zang, Chen Huang, Chen Change Loy

ICCV 2021 · Label-efficient and robust learning › Recognition and long-tail · approach: Decoupled and calibrated long-tail training · 136 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 adaptive feature augmentation (FA) (ablation)Mask R-CNN RFS · LVIS v1.0 valAP: 20.8 -> 22.3Improved
add adaptive feature sampling (FS) on top of FA (ablation)Mask R-CNN RFS + FA · LVIS v1.0 valAP: 22.3 -> 23.7Improved
add adaptive FS on top of FA, rare classes (ablation)Mask R-CNN RFS + FA · LVIS v1.0 valAP rare: 12.7 -> 17.8Improved
use adaptive sampling instead of static s=5 (ablation)FASA · LVIS v1.0 valAP: 22.3 -> 23.7Improved
static scale s=5 vs s=1 on frequent classes (ablation)FASA static sampling · LVIS v1.0 valAP_r and AP_f: s=5 best for AP_r, s=1 best for AP_fMixed
add FASA to vanilla Mask R-CNNMask R-CNN · LVIS v1.0 valAP: 19.3 -> 22.6Improved
add FASA to RFSRFS · LVIS v1.0 valAP: 22.8 -> 24.1Improved
add FASA to EQLEQL · LVIS v1.0 valAP: 22.1 -> 24.4Improved
add FASA to Seesaw lossSeesaw + RFS · LVIS v1.0 valAP: 26.4 -> 27.5Improved
add FASA to Mask R-CNN with ResNet-101Mask R-CNN RFS R101 · LVIS v1.0 valAP: 24.4 -> 26.3Improved
add FASA to Cascade Mask R-CNN with SeesawCascade Mask R-CNN Seesaw R101 · LVIS v1.0 valAP: 30.1 -> 31.5Improved
add FASA to Mask R-CNN on COCO-LTMask R-CNN · COCO-LT minivalAP mask: 18.7 -> 23.4Improved

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