AI Paper Atlas

Research map / Detection and segmentation / CNN features

Region-Based Semantic Segmentation with End-to-End Training

Holger Caesar, Jasper Uijlings, Vittorio Ferrari

ECCV 2016 · Detection and segmentation › CNN features · approach: Boundary-aware instance segmentation · 76 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 region-to-pixel layer with end-to-end training (ablation)Region-based baseline with bounding box pooling · SIFT Flow testglobal accuracy: 60.9 -> 83.7Improved
apply max before softmax (ablation)Softmax-before-max variant · SIFT Flow testclass-average accuracy: +8.7%Improved
use multi-scale overlapping regions instead of non-overlapping (ablation)Non-overlapping regions · SIFT Flow testclass-average accuracy: 60.0 -> 64.0Improved
use free-form ROI pooling instead of bounding box pooling (ablation)Bounding box ROI pooling · SIFT Flow testclass-average accuracy: 62.3 -> 62.8Improved
combine free-form and bounding box pooling with tied weights (ablation)Free-form ROI pooling · SIFT Flow testclass-average accuracy: 62.8 -> 63.4Improved
use separate weights for the two pooling streams (ablation)Tied weights combination · SIFT Flow testclass-average accuracy: 63.4 -> 64.0Improved
use free-form regions instead of purely rectangular regions (ablation)Purely rectangular regions · SIFT Flow testclass-average accuracy: 59.3 -> 64.0Improved
use region-based end-to-end model instead of FCN-16s on boundariesFCN-16s · SIFT Flow test, boundary pixelsclass-average accuracy: 37.9 -> 57.3Improved
use region-based end-to-end model instead of FCN-16s on full imageFCN-16s · SIFT Flow test, full imageclass-average accuracy: 49.3 -> 64.0Improved
use region-based end-to-end model instead of FCN-16s on boundariesFCN-16s · PASCAL Context val, boundary pixelsclass-average accuracy: 34.0 -> 38.9Improved
use unbalanced loss model versus Eigen on global accuracyEigen 2015 · SIFT Flow testglobal accuracy: 86.8 -> 84.3Hurt
use our model instead of Long et al. on global accuracyLong et al. · PASCAL Context valglobal accuracy: 65.9 -> 62.4Hurt

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