AI Paper Atlas

Research map / Detection and segmentation / CNN features

EfficientFCN: Holistically-guided Decoding for Semantic Segmentation

Jianbo Liu, Junjun He, Jiawei Zhang, Jimmy S. Ren, Hongsheng Li

ECCV 2020 · Detection and segmentation › CNN features · approach: Context modeling for semantic segmentation · 67 citations (OpenAlex)

Read the paper

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 OS=16 and OS=32 features to the fused map used for codebook generation (ablation)EfficientFCN · PASCAL ContextmIoU: 54.8 -> 55.1Improved
add OS=8 features as well to the fused map used for codebook generation (ablation)EfficientFCN · PASCAL ContextmIoU: 55.1 -> 55.3Improved
add OS=16 features to the fused map used for codeword assembly coefficients (ablation)EfficientFCN · PASCAL ContextmIoU: 47.9 -> 52.1Improved
add OS=32 features as well to the fused map used for codeword assembly coefficients (ablation)EfficientFCN · PASCAL ContextmIoU: 52.1 -> 55.3Improved
add the average codeword base transfer to the assembly guidance features (ablation)EfficientFCN · PASCAL ContextmIoU: 54.2 -> 55.3Improved
increase number of codewords from 32 to 512 (ablation)EfficientFCN · PASCAL ContextmIoU: 54.5 -> 55.5Improved
increase number of codewords from 512 to 1024 (ablation)EfficientFCN · PASCAL ContextmIoU: 55.5 -> 55.1Hurt
increase number of codewords from 256 to 512 (ablation)EfficientFCN · PASCAL ContextGFlops: 69.6 -> 72.1Hurt
replace bilinear-upsampling U-Net decoder with holistically-guided decoderResNet101 encoder-decoder · PASCAL ContextmIoU: 49.3 -> 55.3Improved
replace dilated ResNet101 FCN with non-dilated ResNet101 plus holistically-guided decoderdilatedFCN-8s · PASCAL ContextmIoU: 47.2 -> 55.3Improved
replace dilated ResNet101 FCN with non-dilated ResNet101 plus holistically-guided decoderdilatedFCN-8s · PASCAL ContextGFlops: 223.6 -> 69.6Improved
add COCO pre-trainingEfficientFCN · PASCAL VOC 2012 testmIoU: 85.4 -> 87.6Improved

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