AI Paper Atlas

Research map / Detection and segmentation / Weakly supervised detection and segmentation

USAGE: A Unified Seed Area Generation Paradigm for Weakly Supervised Semantic Segmentation

Zelin Peng, Guanchun Wang, Lingxi Xie, Dongsheng Jiang, Wei Shen, Qi Tian

ICCV 2023 · Detection and segmentation › Weakly supervised detection and segmentation · approach: CAM refinement and prototypes · 51 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 activation shape controlling (temperature) to CAM seed area (ablation)USAGE · PASCAL VOC train seed (Transformer)mIoU: 52.3 -> 64.0Improved
add activation shape controlling (temperature) to CAM seed area (ablation)USAGE · PASCAL VOC train seed (CNN)mIoU: 47.8 -> 49.6Improved
add activation shape regularization (network adjustment consistency) (ablation)USAGE · PASCAL VOC train seed (CNN)mIoU: 49.6 -> 57.7Improved
add activation shape regularization (ablation)USAGE · PASCAL VOC train seed (Transformer)mIoU: 64.0 -> 67.7Improved
use sharpening (small tau1) for Transformers instead of smoothing (large tau1) (ablation)USAGE · PASCAL VOC train seed (Transformer)mIoU: 50.3 -> 67.7Improved
use smoothing (large tau1=50) for CNNs instead of sharpening (ablation)USAGE · PASCAL VOC train seed (CNN)mIoU: 48.9 -> 57.7Improved
vary temperature tau1 within the correct regime (ablation)USAGE · PASCAL VOC train seedmIoU: 67.2 -> 67.7 -> 67.1No clear effect
replace network adjustment with geometric transform regularization (ablation)USAGE (Transformer) · PASCAL VOC train seedmIoU: 67.7 -> 62.4Hurt
replace network adjustment with online attention accumulation (ablation)USAGE (Transformer) · PASCAL VOC train seedmIoU: 67.7 -> 65.1Hurt
use self-adaptive adjustment gap instead of fixed gap (ablation)USAGE (Transformer) · PASCAL VOC train seedmIoU: 66.6 -> 67.7Improved
use self-adaptive adjustment gap instead of fixed gap (ablation)USAGE (CNN) · PASCAL VOC train seedmIoU: 55.5 -> 57.7Improved
add architecture adjustment regularization to CAM mapping (ablation)CAM (Transformer) · PASCAL VOC train seedmIoU: 52.3 -> 56.5Improved

Later papers that compared with this method

5 later accepted papers report USAGE as a baseline in their results tables; 5 of them report a better number on most of the results they share with it.

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