paper-with-me

Semantic Segmentation 벤치마크

Semantic Segmentation on PASCAL Context

132개 결과 · ⬇ CSV · JSON

mIoU

24.7 36.3 47.9 59.5 71.1 2014-11 2026-11 FCN-8s — 37.8 (2014-11-14) FCN-8s — 37.8 (2014-11-14) CFM — 34.4 (2014-12-03) CFM — 34.4 (2014-12-03) CRF-RNN — 39.3 (2015-02-11) CRF-RNN — 39.3 (2015-02-11) BoxSup — 40.5 (2015-03-05) BoxSup — 40.5 (2015-03-05) Piecewise — 43.3 (2015-04-04) Piecewise — 43.3 (2015-04-04) ParseNet — 40.4 (2015-06-15) ParseNet — 40.4 (2015-06-15) HO CRF — 41.3 (2015-11-25) HO CRF — 41.3 (2015-11-25) VeryDeep — 44.5 (2016-05-23) VeryDeep — 44.5 (2016-05-23) DeepLabV2 — 45.7 (2016-06-02) DeepLabV2 — 45.7 (2016-06-02) RBE2E — 32.5 (2016-07-26) RBE2E — 32.5 (2016-07-26) RefineNet — 47.3 (2016-11-20) RefineNet — 47.3 (2016-11-20) ResNet-38 — 48.1 (2016-11-30) ResNet-38 — 48.1 (2016-11-30) PSPNet (ResNet-101) — 47.8 (2016-12-04) PSPNet (ResNet-101) — 47.8 (2016-12-04) Dilated-FCN2s — 42.6 (2017-07-26) Dilated-FCN2s — 42.6 (2017-07-26) EncNet (ResNet-101) — 51.7 (2018-03-23) EncNet (ResNet-101) — 51.7 (2018-03-23) DANet (ResNet-101) — 52.6 (2018-09-09) DANet (ResNet-101) — 52.6 (2018-09-09) DUpsampling — 52.5 (2019-03-05) DUpsampling — 52.5 (2019-03-05) Joint Pyramid Upsampling + EncNet — 53.1 (2019-03-28) Joint Pyramid Upsampling + EncNet — 53.1 (2019-03-28) CFNet (ResNet-101) — 54.0 (2019-06-01) ICM — 52.6 (2019-06-01) CFNet (ResNet-50) — 51.5 (2019-06-01) CFNet (ResNet-101) — 54.0 (2019-06-01) ICM — 52.6 (2019-06-01) CFNet (ResNet-50) — 51.5 (2019-06-01) EMANet — 53.1 (2019-07-31) EMANet — 53.1 (2019-07-31) CFNet (ResNet-101) — 54.0 (2019-08-20) HRNetV2 HRNetV2-W48 — 54.0 (2019-08-20) CFNet (ResNet-101) — 54.0 (2019-08-20) HRNetV2 HRNetV2-W48 — 54.0 (2019-08-20) Asymmetric ALNN — 52.8 (2019-08-21) Asymmetric ALNN — 52.8 (2019-08-21) BFP — 53.6 (2019-08-31) BFP — 53.6 (2019-08-31) SVCNet (ResNet-101) — 53.2 (2019-09-05) SVCNet (ResNet-101) — 53.2 (2019-09-05) DGCNet (MS, ResNet-101) — 53.7 (2019-09-13) DGCNet (MS, ResNet-101) — 53.7 (2019-09-13) HRNetV2 + OCR + RMI (PaddleClas pretrained) — 59.6 (2019-09-24) OCR (HRNetV2-W48) — 56.2 (2019-09-24) OCR (ResNet-101) — 54.8 (2019-09-24) HRNetV2 + OCR + RMI (PaddleClas pretrained) — 59.6 (2019-09-24) OCR (HRNetV2-W48) — 56.2 (2019-09-24) OCR (ResNet-101) — 54.8 (2019-09-24) LaU-regression-loss (ResNet-101) — 53.9 (2019-11-13) LaU-regression-loss (ResNet-101) — 53.9 (2019-11-13) DCNAS — 55.6 (2020-03-26) DCNAS — 55.6 (2020-03-26) CPN(ResNet-101) — 53.9 (2020-04-03) CPN(ResNet-101) — 53.9 (2020-04-03) ResNeSt-269 — 58.9 (2020-04-19) ResNeSt-200 — 58.4 (2020-04-19) ResNeSt-101 — 56.5 (2020-04-19) ResNeSt-269 — 58.9 (2020-04-19) ResNeSt-200 — 58.4 (2020-04-19) ResNeSt-101 — 56.5 (2020-04-19) DNL — 55.3 (2020-06-11) DNL — 55.3 (2020-06-11) DRAN(ResNet-101) — 55.4 (2020-08-05) DRAN(ResNet-101) — 55.4 (2020-08-05) SETR-MLA (16, 80k, MS) — 55.83 (2020-12-31) SETR-MLA (16, 80k, MS) — 55.83 (2020-12-31) CAA + Simple decoder (Efficientnet-B7) — 60.5 (2021-01-19) CAA (Efficientnet-B7) — 60.1 (2021-01-19) CAA (ResNet-101) — 55.0 (2021-01-19) CAA + Simple decoder (Efficientnet-B7) — 60.5 (2021-01-19) CAA (Efficientnet-B7) — 60.1 (2021-01-19) CAA (ResNet-101) — 55.0 (2021-01-19) DPT-Hybrid — 60.46 (2021-03-24) DPT-Hybrid — 60.46 (2021-03-24) CASSOD — 52.76 (2021-04-29) CASSOD — 52.76 (2021-04-29) Seg-L-Mask/16 — 59.0 (2021-05-12) Seg-L-Mask/16 — 59.0 (2021-05-12) HamNet (ResNet-101) — 55.2 (2021-09-09) HamNet (ResNet-101) — 55.2 (2021-09-09) CondNet(ResNest-101) — 57.0 (2021-09-21) CondNet(ResNet-101) — 56.0 (2021-09-21) CondNet(ResNest-101) — 57.0 (2021-09-21) CondNet(ResNet-101) — 56.0 (2021-09-21) SenFormer (Swin-L) — 64.0 (2021-11-26) SenFormer (ResNet-101) — 56.6 (2021-11-26) SenFormer (Swin-L) — 64.0 (2021-11-26) SenFormer (ResNet-101) — 56.6 (2021-11-26) SIW(Segformer-B5) — 54.2 (2022-02-04) SIW(Segformer-B5) — 54.2 (2022-02-04) CAA + CAR (ConvNeXt-Large + JPU) — 64.1 (2022-03-14) CAA + CAR (ConvNeXt-Large + JPU) — 64.1 (2022-03-14) ViT-Adapter-L (Mask2Former, BEiT pretrain) — 68.2 (2022-05-17) ViT-Adapter-L (UperNet, BEiT pretrain) — 67.5 (2022-05-17) ViT-Adapter-L (Mask2Former, BEiT pretrain) — 68.2 (2022-05-17) ViT-Adapter-L (UperNet, BEiT pretrain) — 67.5 (2022-05-17) GPaCo (ResNet101) — 56.2 (2022-09-26) GPaCo (ResNet101) — 56.2 (2022-09-26) Sequential Ensemble (Segformer + HRNet) — 62.1 (2022-10-08) Sequential Ensemble (Segformer + HRNet) — 62.1 (2022-10-08) SegViT (ours) — 65.3 (2022-10-12) SegViT (ours) — 65.3 (2022-10-12) InternImage-H — 70.3 (2022-11-10) InternImage-H — 70.3 (2022-11-10) SegCLIP — 24.7 (2022-11-27) SegCLIP — 24.7 (2022-11-27) RSSeg-ViT-L (BEiT pretrain) — 68.9 (2022-12-28) RSSeg-ViT-L — 67.5 (2022-12-28) RSSeg-ViT-L (BEiT pretrain) — 68.9 (2022-12-28) RSSeg-ViT-L — 67.5 (2022-12-28) PlainSeg (EVA-02-L) — 71.0 (2023-10-19) PlainSeg (EVA-02-L) — 71.0 (2023-10-19) DEPICT-SA (ViT-L multi-scale) — 58.6 (2024-11-05) DEPICT-SA (ViT-L single-scale) — 57.9 (2024-11-05) DEPICT-SA (ViT-L multi-scale) — 58.6 (2024-11-05) DEPICT-SA (ViT-L single-scale) — 57.9 (2024-11-05) VPNeXt — 71.1 (2025-02-23) VPNeXt — 71.1 (2025-02-23) FCN-8s — 37.8 (2014-11-14) CRF-RNN — 39.3 (2015-02-11) BoxSup — 40.5 (2015-03-05) Piecewise — 43.3 (2015-04-04) VeryDeep — 44.5 (2016-05-23) DeepLabV2 — 45.7 (2016-06-02) RefineNet — 47.3 (2016-11-20) ResNet-38 — 48.1 (2016-11-30) EncNet (ResNet-101) — 51.7 (2018-03-23) DANet (ResNet-101) — 52.6 (2018-09-09) Joint Pyramid Upsampling + EncNet — 53.1 (2019-03-28) CFNet (ResNet-101) — 54.0 (2019-06-01) HRNetV2 + OCR + RMI (PaddleClas pretrained) — 59.6 (2019-09-24) CAA + Simple decoder (Efficientnet-B7) — 60.5 (2021-01-19) SenFormer (Swin-L) — 64.0 (2021-11-26) CAA + CAR (ConvNeXt-Large + JPU) — 64.1 (2022-03-14) ViT-Adapter-L (Mask2Former, BEiT pretrain) — 68.2 (2022-05-17) InternImage-H — 70.3 (2022-11-10) PlainSeg (EVA-02-L) — 71.0 (2023-10-19) VPNeXt — 71.1 (2025-02-23)
RankModel mIoUMean AccuracyPixel Accuracy PaperCodeYear
1 VPNeXt 71.1–– VPNeXt -- Rethinking Dense Decoding for Plain Vision Transformer 2025
2 PlainSeg (EVA-02-L) 71.0–– Minimalist and High-Performance Semantic Segmentation with Plain Vision Transformers ydhonghit/plainseg 2023
3 InternImage-H 70.3–– InternImage: Exploring Large-Scale Vision Foundation Models with Deformable Convolutions opengvlab/internimage · OpenGVLab/M3I-Pretraining · chenller/mmseg-extension 2022
4 RSSeg-ViT-L (BEiT pretrain) 68.9–– Representation Separation for Semantic Segmentation with Vision Transformers 2022
5 ViT-Adapter-L (Mask2Former, BEiT pretrain) 68.2–– Vision Transformer Adapter for Dense Predictions czczup/vit-adapter · chenller/mmseg-extension 2022
6 ViT-Adapter-L (UperNet, BEiT pretrain) 67.5–– Vision Transformer Adapter for Dense Predictions czczup/vit-adapter · chenller/mmseg-extension 2022
6 RSSeg-ViT-L 67.5–– Representation Separation for Semantic Segmentation with Vision Transformers 2022
8 SegViT (ours) 65.3–– SegViT: Semantic Segmentation with Plain Vision Transformers zbwxp/SegVit 2022
9 CAA + CAR (ConvNeXt-Large + JPU) 64.1–– CAR: Class-aware Regularizations for Semantic Segmentation edwardyehuang/CAR 2022
10 SenFormer (Swin-L) 64.0–– Efficient Self-Ensemble for Semantic Segmentation WalBouss/SenFormer 2021
11 Sequential Ensemble (Segformer + HRNet) 62.1–– Sequential Ensembling for Semantic Segmentation 2022
12 CAA + Simple decoder (Efficientnet-B7) 60.5–– Channelized Axial Attention for Semantic Segmentation -- Considering Channel Relation within Spatial Attention for Semantic Segmentation edwardyehuang/CAA 2021
13 DPT-Hybrid 60.46–– Vision Transformers for Dense Prediction huggingface/transformers · isl-org/MiDaS · intel-isl/MiDaS · +12 2021
14 CAA (Efficientnet-B7) 60.1–– Channelized Axial Attention for Semantic Segmentation -- Considering Channel Relation within Spatial Attention for Semantic Segmentation edwardyehuang/CAA 2021
15 HRNetV2 + OCR + RMI (PaddleClas pretrained) 59.6–– Segmentation Transformer: Object-Contextual Representations for Semantic Segmentation open-mmlab/mmsegmentation · PaddlePaddle/PaddleSeg · HRNet/HRNet-Semantic-Segmentation · +8 2019
16 Seg-L-Mask/16 59.0–– Segmenter: Transformer for Semantic Segmentation PaddlePaddle/PaddleSeg · BR-IDL/PaddleViT · rstrudel/segmenter · +5 2021
17 ResNeSt-269 58.9–– ResNeSt: Split-Attention Networks rwightman/pytorch-image-models · open-mmlab/mmdetection · open-mmlab/mmpose · +33 2020
18 DEPICT-SA (ViT-L multi-scale) 58.6–– Rethinking Decoders for Transformer-based Semantic Segmentation: A Compression Perspective qishuaiwen/depict 2024
19 ResNeSt-200 58.4–– ResNeSt: Split-Attention Networks rwightman/pytorch-image-models · open-mmlab/mmdetection · open-mmlab/mmpose · +33 2020
20 DEPICT-SA (ViT-L single-scale) 57.9–– Rethinking Decoders for Transformer-based Semantic Segmentation: A Compression Perspective qishuaiwen/depict 2024
1–20 / 132 다음 → 페이지당 10 20 50 100