Papers › ICNet for Real-Time Semantic Segmentation on High-Resolution Images

ICNet for Real-Time Semantic Segmentation on High-Resolution Images

27 Apr 2017ECCV 2018 9arXiv:1704.08545archive 2025-07-28

Hengshuang Zhao, Xiaojuan Qi, Xiaoyong Shen, Jianping Shi, Jiaya Jia

We focus on the challenging task of real-time semantic segmentation in this paper. It finds many practical applications and yet is with fundamental difficulty of reducing a large portion of computation for pixel-wise label inference. We propose an image cascade network (ICNet) that incorporates multi-resolution branches under proper label guidance to address this challenge. We provide in-depth analysis of our framework and introduce the cascade feature fusion unit to quickly achieve high-quality segmentation. Our system yields real-time inference on a single GPU card with decent quality results evaluated on challenging datasets like Cityscapes, CamVid and COCO-Stuff.

PaperPDFConference PDFCode

In Syntology View this paper on Syntology, its page in Syntology's graph. That page lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, on Syntology's MCP service (how to connect):

Code

18 repositories listed; official and paper-mentioned ones first.

hszhao/ICNet officialmentioned in paperSyntology: not harvested report
Bigpingping97/ICNet mentioned on GitHubmindsporeSyntology: not harvested report
GuangyanZhang/SCNN-Deeplabv3-bisenet-icnet mentioned on GitHubpaddleSyntology: not harvested report
Mind23-2/MindCode-53 mentioned on GitHubmindsporeSyntology: not harvested report
daheyinyin/ICNet mentioned on GitHubmindsporeSyntology: not harvested report
hellochick/ICNet-tensorflow mentioned on GitHubtfSyntology: not harvested report
liminn/icnet-pytorch mentioned on GitHubpytorchSyntology: not harvested report
lisilin013/ICNet-tensorflow-ros mentioned on GitHubtfSyntology: not harvested report
lyqcom/icnet mentioned on GitHubmindsporeSyntology: not harvested report
mattangus/fast-semantic-segmentation mentioned on GitHubtfSyntology: not harvested report
oandrienko/fast-semantic-segmentation mentioned on GitHubtfSyntology: not harvested report
osmr/imgclsmob mentioned on GitHubmxnetMITSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to ResNeSt: Split-Attention Networks (arXiv:2004.08955; that paper's own run record: 28 ran (of which 0 constructed an object rather than computing a result; 21 with no instrument failure: 0 honoured, 0 violated, 21 with no contract checked; 7 where Syntology's instrument failed) · 20 unverified); nothing checks that they implement this paper's method. report
pooruss/ICNet-Paddle2.2.0rc mentioned on GitHubpaddleSyntology: not harvested report
victorpham1997/Local_ICNet mentioned on GitHubtfSyntology: not harvested report
zh320/realtime-semantic-segmentation-pytorch mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to CGNet: A Light-weight Context Guided Network for Semantic Segmentation (arXiv:1811.08201; that paper's own run record: 2 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 1 violated, 0 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified), ESPNetv2: A Light-weight, Power Efficient, and General Purpose Convolutional Neural Network (arXiv:1811.11431; that paper's own run record: 4 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 2 where Syntology's instrument failed) · 0 unverified), Lite-HRNet: A Lightweight High-Resolution Network (arXiv:2104.06403; that paper's own run record: 13 ran (of which 2 constructed an object rather than computing a result; 11 with no instrument failure: 0 honoured, 0 violated, 11 with no contract checked; 2 where Syntology's instrument failed) · 16 unverified) and 1 more (every one in this page's JSON twin); nothing checks that they implement this paper's method. report
Mind23-2/MindCode-3 mindsporeSyntology: not harvested report
mindspore-ai/models mindsporeSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to AdaBin: Improving Binary Neural Networks with Adaptive Binary Sets (arXiv:2208.08084; that paper's own run record: 4 ran (of which 0 constructed an object rather than computing a result; 4 with no instrument failure: 0 honoured, 0 violated, 4 with no contract checked; 0 where Syntology's instrument failed) · 1 unverified); nothing checks that they implement this paper's method. report
yfjcode/ICNet mindsporeSyntology: not harvested report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

A paper named beside a repository with no sample linked to this paper is shown with that paper's own run record, not this paper's: “ran” means executed on a synthesized input, not that the code is correct, and “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Papers are listed in arXiv-id order, at most three per repository.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Dichotomous Image SegmentationReal-Time Semantic SegmentationSegmentationSemantic SegmentationThermal Image SegmentationVocal Bursts Intensity Prediction

1 archive task tag without a task page not shown.

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Dichotomous Image Segmentation DIS-TE1 ICNet E-measure 0.784 #17 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE1 ICNet HCE 234 #17 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE1 ICNet MAE 0.095 #17 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE1 ICNet S-Measure 0.716 #17 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE1 ICNet max F-Measure 0.631 #17 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE1 ICNet weighted F-measure 0.535 #17 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE2 ICNet E-measure 0.826 #15 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE2 ICNet HCE 512 #15 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE2 ICNet MAE 0.095 #15 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE2 ICNet S-Measure 0.759 #15 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE2 ICNet max F-Measure 0.716 #15 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE2 ICNet weighted F-measure 0.627 #15 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE3 ICNet E-measure 0.852 #12 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE3 ICNet HCE 1001 #12 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE3 ICNet MAE 0.091 #12 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE3 ICNet S-Measure 0.780 #12 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE3 ICNet max F-Measure 0.752 #12 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE3 ICNet weighted F-measure 0.664 #12 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE4 ICNet E-measure 0.837 #12 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE4 ICNet HCE 3690 #12 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE4 ICNet MAE 0.099 #12 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE4 ICNet S-Measure 0.776 #12 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE4 ICNet max F-Measure 0.749 #12 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-TE4 ICNet weighted F-measure 0.663 #12 of 22 Archive leaderboard report
Dichotomous Image Segmentation DIS-VD ICNet E-measure 0.811 #14 of 24 Archive leaderboard report
Dichotomous Image Segmentation DIS-VD ICNet HCE 1503 #14 of 24 Archive leaderboard report
Dichotomous Image Segmentation DIS-VD ICNet MAE 0.102 #14 of 24 Archive leaderboard report
Dichotomous Image Segmentation DIS-VD ICNet S-Measure 0.747 #14 of 24 Archive leaderboard report
Dichotomous Image Segmentation DIS-VD ICNet max F-Measure 0.697 #14 of 24 Archive leaderboard report
Dichotomous Image Segmentation DIS-VD ICNet weighted F-measure 0.609 #14 of 24 Archive leaderboard report
Real-Time Semantic Segmentation CamVid ICNet Frame (fps) 27.8 #23 of 29 Archive leaderboard report
Real-Time Semantic Segmentation CamVid ICNet Time (ms) 36 #23 of 29 Archive leaderboard report
Real-Time Semantic Segmentation CamVid ICNet mIoU 67.1% #23 of 29 Archive leaderboard report
Real-Time Semantic Segmentation Cityscapes test ICNet Frame (fps) 30.3 #31 of 39 Archive leaderboard report
Real-Time Semantic Segmentation Cityscapes test ICNet Time (ms) 33 #31 of 39 Archive leaderboard report
Real-Time Semantic Segmentation Cityscapes test ICNet mIoU 70.6% #31 of 39 Archive leaderboard report
Semantic Segmentation BDD100K val ICNet mIoU 52.4(39.5fps) #13 of 24 Archive leaderboard report
Semantic Segmentation Cityscapes test ICNet Mean IoU (class) 70.6% #79 of 105 Archive leaderboard report
Semantic Segmentation Trans10K ICNet GFLOPs 10.64 #14 of 15 Archive leaderboard report
Semantic Segmentation Trans10K ICNet mIoU 23.39% #14 of 15 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

1x1 ConvolutionBatch NormalizationDilated ConvolutionSoftmax

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections