TasksSotADatasetsPapersMethodsSubmitAbout
Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable BenchmarksAll SotADatasetsPapersMethods

Community

Submit ResultsAbout

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Papers/Pyramid Scene Parsing Network

Pyramid Scene Parsing Network

Hengshuang Zhao, Jianping Shi, Xiaojuan Qi, Xiaogang Wang, Jiaya Jia

2016-12-04CVPR 2017 7Scene ParsingThermal Image SegmentationDichotomous Image SegmentationImage ClassificationReal-Time Semantic SegmentationLesion SegmentationSemantic SegmentationVideo Semantic Segmentation
PaperPDFCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCode(official)CodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCodeCode

Abstract

Scene parsing is challenging for unrestricted open vocabulary and diverse scenes. In this paper, we exploit the capability of global context information by different-region-based context aggregation through our pyramid pooling module together with the proposed pyramid scene parsing network (PSPNet). Our global prior representation is effective to produce good quality results on the scene parsing task, while PSPNet provides a superior framework for pixel-level prediction tasks. The proposed approach achieves state-of-the-art performance on various datasets. It came first in ImageNet scene parsing challenge 2016, PASCAL VOC 2012 benchmark and Cityscapes benchmark. A single PSPNet yields new record of mIoU accuracy 85.4% on PASCAL VOC 2012 and accuracy 80.2% on Cityscapes.

Results

TaskDatasetMetricValueModel
Medical Image SegmentationAnatomical Tracings of Lesions After Stroke (ATLAS)Dice0.3571PSPNet
Medical Image SegmentationAnatomical Tracings of Lesions After Stroke (ATLAS)IoU0.254PSPNet
Medical Image SegmentationAnatomical Tracings of Lesions After Stroke (ATLAS)Precision0.4769PSPNet
Medical Image SegmentationAnatomical Tracings of Lesions After Stroke (ATLAS)Recall0.3335PSPNet
Scene ParsingCityscapes valmIoU79.7PSPNet-101 [20]
Scene ParsingCityscapes valmIoU78.1PSPNet-50 [20]
Scene ParsingCamVidMean IoU76PSPNet-50
Semantic Segmentation US3DmIoU73.12PSNet
Semantic SegmentationFine-Grained Grass Segmentation DatasetmIoU47.95PSPNet
Semantic SegmentationCityscapes valmIoU79.7PSPNet (Dilated-ResNet-101)
Semantic SegmentationBDD100K valmIoU62.3PSPNet
Semantic SegmentationSELMAmIoU68.4PSPNet
Semantic Segmentation PotsdammIoU82.98PSPNet
Semantic SegmentationPASCAL ContextmIoU47.8PSPNet (ResNet-101)
Semantic SegmentationUrbanLFmIoU (Real)76.34PSPNet
Semantic SegmentationUrbanLFmIoU (Syn)75.78PSPNet
Semantic SegmentationVaihingenmIoU76.79PSPNet
Semantic SegmentationTrans10KGFLOPs187.03PSPNet
Semantic SegmentationDADA-segmIoU20.1PSPNet (ResNet-101)
Semantic SegmentationADE20KTest Score55.38PSPNet
Semantic SegmentationADE20KValidation mIoU44.94PSPNet
Semantic SegmentationADE20KValidation mIoU43.51PSPNet (ResNet-152)
Semantic SegmentationADE20KValidation mIoU43.29PSPNet (ResNet-101)
Semantic SegmentationMFN DatasetmIOU46.1PSPNet
Semantic SegmentationCamVidFrame (fps)5.4PSPNet
Semantic SegmentationCamVidTime (ms)185PSPNet
Semantic SegmentationNYU Depth v2Speed(ms/f)72PSPNet101
Semantic SegmentationNYU Depth v2mIoU43.2PSPNet101
Semantic SegmentationNYU Depth v2Speed(ms/f)47PSPNet50
Semantic SegmentationNYU Depth v2mIoU41.8PSPNet50
Semantic SegmentationNYU Depth v2Speed(ms/f)19PSPNet18
Semantic SegmentationNYU Depth v2mIoU35.9PSPNet18
Object DetectionDIS-TE4E-measure0.815PSPNet
Object DetectionDIS-TE4HCE3806PSPNet
Object DetectionDIS-TE4MAE0.107PSPNet
Object DetectionDIS-TE4S-Measure0.758PSPNet
Object DetectionDIS-TE4max F-Measure0.725PSPNet
Object DetectionDIS-TE4weighted F-measure0.63PSPNet
Object DetectionDIS-VDE-measure0.802PSPNet
Object DetectionDIS-VDHCE1588PSPNet
Object DetectionDIS-VDMAE0.102PSPNet
Object DetectionDIS-VDS-Measure0.744PSPNet
Object DetectionDIS-VDmax F-Measure0.691PSPNet
Object DetectionDIS-VDweighted F-measure0.603PSPNet
Object DetectionDIS-TE2E-measure0.828PSPNet
Object DetectionDIS-TE2HCE586PSPNet
Object DetectionDIS-TE2MAE0.092PSPNet
Object DetectionDIS-TE2S-Measure0.763PSPNet
Object DetectionDIS-TE2max F-Measure0.724PSPNet
Object DetectionDIS-TE2weighted F-measure0.636PSPNet
Object DetectionDIS-TE1E-measure0.791PSPNet
Object DetectionDIS-TE1HCE267PSPNet
Object DetectionDIS-TE1MAE0.089PSPNet
Object DetectionDIS-TE1S-Measure0.725PSPNet
Object DetectionDIS-TE1max F-Measure0.645PSPNet
Object DetectionDIS-TE1weighted F-measure0.557PSPNet
Object DetectionDIS-TE3E-measure0.843PSPNet
Object DetectionDIS-TE3HCE1111PSPNet
Object DetectionDIS-TE3MAE0.092PSPNet
Object DetectionDIS-TE3S-Measure0.774PSPNet
Object DetectionDIS-TE3max F-Measure0.747PSPNet
Object DetectionDIS-TE3weighted F-measure0.657PSPNet
3DDIS-TE4E-measure0.815PSPNet
3DDIS-TE4HCE3806PSPNet
3DDIS-TE4MAE0.107PSPNet
3DDIS-TE4S-Measure0.758PSPNet
3DDIS-TE4max F-Measure0.725PSPNet
3DDIS-TE4weighted F-measure0.63PSPNet
3DDIS-VDE-measure0.802PSPNet
3DDIS-VDHCE1588PSPNet
3DDIS-VDMAE0.102PSPNet
3DDIS-VDS-Measure0.744PSPNet
3DDIS-VDmax F-Measure0.691PSPNet
3DDIS-VDweighted F-measure0.603PSPNet
3DDIS-TE2E-measure0.828PSPNet
3DDIS-TE2HCE586PSPNet
3DDIS-TE2MAE0.092PSPNet
3DDIS-TE2S-Measure0.763PSPNet
3DDIS-TE2max F-Measure0.724PSPNet
3DDIS-TE2weighted F-measure0.636PSPNet
3DDIS-TE1E-measure0.791PSPNet
3DDIS-TE1HCE267PSPNet
3DDIS-TE1MAE0.089PSPNet
3DDIS-TE1S-Measure0.725PSPNet
3DDIS-TE1max F-Measure0.645PSPNet
3DDIS-TE1weighted F-measure0.557PSPNet
3DDIS-TE3E-measure0.843PSPNet
3DDIS-TE3HCE1111PSPNet
3DDIS-TE3MAE0.092PSPNet
3DDIS-TE3S-Measure0.774PSPNet
3DDIS-TE3max F-Measure0.747PSPNet
3DDIS-TE3weighted F-measure0.657PSPNet
Video Semantic SegmentationCityscapes valmIoU79.7PSPNet-101 [20]
Video Semantic SegmentationCityscapes valmIoU78.1PSPNet-50 [20]
Video Semantic SegmentationCamVidMean IoU76PSPNet-50
Scene UnderstandingCityscapes valmIoU79.7PSPNet-101 [20]
Scene UnderstandingCityscapes valmIoU78.1PSPNet-50 [20]
Scene UnderstandingCamVidMean IoU76PSPNet-50
RGB Salient Object DetectionDIS-TE4E-measure0.815PSPNet
RGB Salient Object DetectionDIS-TE4HCE3806PSPNet
RGB Salient Object DetectionDIS-TE4MAE0.107PSPNet
RGB Salient Object DetectionDIS-TE4S-Measure0.758PSPNet
RGB Salient Object DetectionDIS-TE4max F-Measure0.725PSPNet
RGB Salient Object DetectionDIS-TE4weighted F-measure0.63PSPNet
RGB Salient Object DetectionDIS-VDE-measure0.802PSPNet
RGB Salient Object DetectionDIS-VDHCE1588PSPNet
RGB Salient Object DetectionDIS-VDMAE0.102PSPNet
RGB Salient Object DetectionDIS-VDS-Measure0.744PSPNet
RGB Salient Object DetectionDIS-VDmax F-Measure0.691PSPNet
RGB Salient Object DetectionDIS-VDweighted F-measure0.603PSPNet
RGB Salient Object DetectionDIS-TE2E-measure0.828PSPNet
RGB Salient Object DetectionDIS-TE2HCE586PSPNet
RGB Salient Object DetectionDIS-TE2MAE0.092PSPNet
RGB Salient Object DetectionDIS-TE2S-Measure0.763PSPNet
RGB Salient Object DetectionDIS-TE2max F-Measure0.724PSPNet
RGB Salient Object DetectionDIS-TE2weighted F-measure0.636PSPNet
RGB Salient Object DetectionDIS-TE1E-measure0.791PSPNet
RGB Salient Object DetectionDIS-TE1HCE267PSPNet
RGB Salient Object DetectionDIS-TE1MAE0.089PSPNet
RGB Salient Object DetectionDIS-TE1S-Measure0.725PSPNet
RGB Salient Object DetectionDIS-TE1max F-Measure0.645PSPNet
RGB Salient Object DetectionDIS-TE1weighted F-measure0.557PSPNet
RGB Salient Object DetectionDIS-TE3E-measure0.843PSPNet
RGB Salient Object DetectionDIS-TE3HCE1111PSPNet
RGB Salient Object DetectionDIS-TE3MAE0.092PSPNet
RGB Salient Object DetectionDIS-TE3S-Measure0.774PSPNet
RGB Salient Object DetectionDIS-TE3max F-Measure0.747PSPNet
RGB Salient Object DetectionDIS-TE3weighted F-measure0.657PSPNet
2D Semantic SegmentationCityscapes valmIoU79.7PSPNet-101 [20]
2D Semantic SegmentationCityscapes valmIoU78.1PSPNet-50 [20]
2D Semantic SegmentationCamVidMean IoU76PSPNet-50
2D ClassificationDIS-TE4E-measure0.815PSPNet
2D ClassificationDIS-TE4HCE3806PSPNet
2D ClassificationDIS-TE4MAE0.107PSPNet
2D ClassificationDIS-TE4S-Measure0.758PSPNet
2D ClassificationDIS-TE4max F-Measure0.725PSPNet
2D ClassificationDIS-TE4weighted F-measure0.63PSPNet
2D ClassificationDIS-VDE-measure0.802PSPNet
2D ClassificationDIS-VDHCE1588PSPNet
2D ClassificationDIS-VDMAE0.102PSPNet
2D ClassificationDIS-VDS-Measure0.744PSPNet
2D ClassificationDIS-VDmax F-Measure0.691PSPNet
2D ClassificationDIS-VDweighted F-measure0.603PSPNet
2D ClassificationDIS-TE2E-measure0.828PSPNet
2D ClassificationDIS-TE2HCE586PSPNet
2D ClassificationDIS-TE2MAE0.092PSPNet
2D ClassificationDIS-TE2S-Measure0.763PSPNet
2D ClassificationDIS-TE2max F-Measure0.724PSPNet
2D ClassificationDIS-TE2weighted F-measure0.636PSPNet
2D ClassificationDIS-TE1E-measure0.791PSPNet
2D ClassificationDIS-TE1HCE267PSPNet
2D ClassificationDIS-TE1MAE0.089PSPNet
2D ClassificationDIS-TE1S-Measure0.725PSPNet
2D ClassificationDIS-TE1max F-Measure0.645PSPNet
2D ClassificationDIS-TE1weighted F-measure0.557PSPNet
2D ClassificationDIS-TE3E-measure0.843PSPNet
2D ClassificationDIS-TE3HCE1111PSPNet
2D ClassificationDIS-TE3MAE0.092PSPNet
2D ClassificationDIS-TE3S-Measure0.774PSPNet
2D ClassificationDIS-TE3max F-Measure0.747PSPNet
2D ClassificationDIS-TE3weighted F-measure0.657PSPNet
Scene SegmentationMFN DatasetmIOU46.1PSPNet
2D Object DetectionDIS-TE4E-measure0.815PSPNet
2D Object DetectionDIS-TE4HCE3806PSPNet
2D Object DetectionDIS-TE4MAE0.107PSPNet
2D Object DetectionDIS-TE4S-Measure0.758PSPNet
2D Object DetectionDIS-TE4max F-Measure0.725PSPNet
2D Object DetectionDIS-TE4weighted F-measure0.63PSPNet
2D Object DetectionDIS-VDE-measure0.802PSPNet
2D Object DetectionDIS-VDHCE1588PSPNet
2D Object DetectionDIS-VDMAE0.102PSPNet
2D Object DetectionDIS-VDS-Measure0.744PSPNet
2D Object DetectionDIS-VDmax F-Measure0.691PSPNet
2D Object DetectionDIS-VDweighted F-measure0.603PSPNet
2D Object DetectionDIS-TE2E-measure0.828PSPNet
2D Object DetectionDIS-TE2HCE586PSPNet
2D Object DetectionDIS-TE2MAE0.092PSPNet
2D Object DetectionDIS-TE2S-Measure0.763PSPNet
2D Object DetectionDIS-TE2max F-Measure0.724PSPNet
2D Object DetectionDIS-TE2weighted F-measure0.636PSPNet
2D Object DetectionDIS-TE1E-measure0.791PSPNet
2D Object DetectionDIS-TE1HCE267PSPNet
2D Object DetectionDIS-TE1MAE0.089PSPNet
2D Object DetectionDIS-TE1S-Measure0.725PSPNet
2D Object DetectionDIS-TE1max F-Measure0.645PSPNet
2D Object DetectionDIS-TE1weighted F-measure0.557PSPNet
2D Object DetectionDIS-TE3E-measure0.843PSPNet
2D Object DetectionDIS-TE3HCE1111PSPNet
2D Object DetectionDIS-TE3MAE0.092PSPNet
2D Object DetectionDIS-TE3S-Measure0.774PSPNet
2D Object DetectionDIS-TE3max F-Measure0.747PSPNet
2D Object DetectionDIS-TE3weighted F-measure0.657PSPNet
2D Object DetectionMFN DatasetmIOU46.1PSPNet
10-shot image generation US3DmIoU73.12PSNet
10-shot image generationFine-Grained Grass Segmentation DatasetmIoU47.95PSPNet
10-shot image generationCityscapes valmIoU79.7PSPNet (Dilated-ResNet-101)
10-shot image generationBDD100K valmIoU62.3PSPNet
10-shot image generationSELMAmIoU68.4PSPNet
10-shot image generation PotsdammIoU82.98PSPNet
10-shot image generationPASCAL ContextmIoU47.8PSPNet (ResNet-101)
10-shot image generationUrbanLFmIoU (Real)76.34PSPNet
10-shot image generationUrbanLFmIoU (Syn)75.78PSPNet
10-shot image generationVaihingenmIoU76.79PSPNet
10-shot image generationTrans10KGFLOPs187.03PSPNet
10-shot image generationDADA-segmIoU20.1PSPNet (ResNet-101)
10-shot image generationADE20KTest Score55.38PSPNet
10-shot image generationADE20KValidation mIoU44.94PSPNet
10-shot image generationADE20KValidation mIoU43.51PSPNet (ResNet-152)
10-shot image generationADE20KValidation mIoU43.29PSPNet (ResNet-101)
10-shot image generationMFN DatasetmIOU46.1PSPNet
10-shot image generationCamVidFrame (fps)5.4PSPNet
10-shot image generationCamVidTime (ms)185PSPNet
10-shot image generationNYU Depth v2Speed(ms/f)72PSPNet101
10-shot image generationNYU Depth v2mIoU43.2PSPNet101
10-shot image generationNYU Depth v2Speed(ms/f)47PSPNet50
10-shot image generationNYU Depth v2mIoU41.8PSPNet50
10-shot image generationNYU Depth v2Speed(ms/f)19PSPNet18
10-shot image generationNYU Depth v2mIoU35.9PSPNet18
16kDIS-TE4E-measure0.815PSPNet
16kDIS-TE4HCE3806PSPNet
16kDIS-TE4MAE0.107PSPNet
16kDIS-TE4S-Measure0.758PSPNet
16kDIS-TE4max F-Measure0.725PSPNet
16kDIS-TE4weighted F-measure0.63PSPNet
16kDIS-VDE-measure0.802PSPNet
16kDIS-VDHCE1588PSPNet
16kDIS-VDMAE0.102PSPNet
16kDIS-VDS-Measure0.744PSPNet
16kDIS-VDmax F-Measure0.691PSPNet
16kDIS-VDweighted F-measure0.603PSPNet
16kDIS-TE2E-measure0.828PSPNet
16kDIS-TE2HCE586PSPNet
16kDIS-TE2MAE0.092PSPNet
16kDIS-TE2S-Measure0.763PSPNet
16kDIS-TE2max F-Measure0.724PSPNet
16kDIS-TE2weighted F-measure0.636PSPNet
16kDIS-TE1E-measure0.791PSPNet
16kDIS-TE1HCE267PSPNet
16kDIS-TE1MAE0.089PSPNet
16kDIS-TE1S-Measure0.725PSPNet
16kDIS-TE1max F-Measure0.645PSPNet
16kDIS-TE1weighted F-measure0.557PSPNet
16kDIS-TE3E-measure0.843PSPNet
16kDIS-TE3HCE1111PSPNet
16kDIS-TE3MAE0.092PSPNet
16kDIS-TE3S-Measure0.774PSPNet
16kDIS-TE3max F-Measure0.747PSPNet
16kDIS-TE3weighted F-measure0.657PSPNet

Related Papers

SeC: Advancing Complex Video Object Segmentation via Progressive Concept Construction2025-07-21Automatic Classification and Segmentation of Tunnel Cracks Based on Deep Learning and Visual Explanations2025-07-18Adversarial attacks to image classification systems using evolutionary algorithms2025-07-17Efficient Adaptation of Pre-trained Vision Transformer underpinned by Approximately Orthogonal Fine-Tuning Strategy2025-07-17Federated Learning for Commercial Image Sources2025-07-17MUPAX: Multidimensional Problem Agnostic eXplainable AI2025-07-17DiffOSeg: Omni Medical Image Segmentation via Multi-Expert Collaboration Diffusion Model2025-07-17SCORE: Scene Context Matters in Open-Vocabulary Remote Sensing Instance Segmentation2025-07-17