AI Paper Atlas

Research map / Detection and segmentation / Open-set and few-shot detection

Handwritten and Printed Text Segmentation: A Signature Case Study

Sina Gholamian, Ali Vahdat

ICCV 2023 · Detection and segmentation › Open-set and few-shot detection · approach: Open-vocabulary semantic segmentation · 13 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
use four-class formulation with overlap class instead of three-class (ablation)FCN-light, Weighted CE · WGM-SYNmean IoU (%): 50 -> 54.02Improved
use four-class formulation with overlap class instead of three-class (ablation)FCN-based, Weighted CE · SignaTR6Kmean IoU (%): 81.09 -> 83.15Improved
add dense CRF post-processing (ablation)segmentation models · SignaTR6Kmean IoU (%): 84.05 -> 81.17 (4-class FCN, CE)Hurt
add CRF heuristic that only relabels background pixels (ablation)4-class FCN-based, CE · SignaTR6KHT IoU (%): 89.21 -> 91.81Improved
add Fine Feature Path in parallel with SSP (ablation)SSP VGG16 · WGM-SYN and SignaTR6Kmean IoU (%): about 7% on WGM-SYN and 8% on SignaTR6KImproved
use VGG16 backbone instead of small FCN-light (ablation)SSP · WGM-SYNmean IoU (%): 54.08 -> 47.17Hurt
use Fusion loss instead of Dice or Weighted Dice on MFM-InceptionV3 (ablation)MFM-InceptionV3 · WGM-SYNmean IoU (%): 40.57 -> 58.77 (Dice vs Fusion)Improved
use Fusion loss instead of CE on MFM-InceptionV3 (ablation)MFM-InceptionV3 · SignaTR6Kmean IoU (%): 88.18 vs 88.15 (CE vs Fusion)No clear effect
best design vs best prior workMFM with CRFH · WGM-SYNmean IoU (%): 50 -> 58.93Improved
best design vs best prior workMFM with CRFH · SignaTR6Kmean IoU (%): 83.02 -> 89.10Improved
use multi-label sigmoid formulation instead of four-class softmax (ablation)segmentation models · text segmentationIoUHurt

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