Papers › CSPNet: A New Backbone that can Enhance Learning Capability of CNN

CSPNet: A New Backbone that can Enhance Learning Capability of CNN

27 Nov 2019arXiv:1911.11929archive 2025-07-28

Chien-Yao Wang, Hong-Yuan Mark Liao, I-Hau Yeh, Yueh-Hua Wu, Ping-Yang Chen, Jun-Wei Hsieh

Neural networks have enabled state-of-the-art approaches to achieve incredible results on computer vision tasks such as object detection. However, such success greatly relies on costly computation resources, which hinders people with cheap devices from appreciating the advanced technology. In this paper, we propose Cross Stage Partial Network (CSPNet) to mitigate the problem that previous works require heavy inference computations from the network architecture perspective. We attribute the problem to the duplicate gradient information within network optimization. The proposed networks respect the variability of the gradients by integrating feature maps from the beginning and the end of a network stage, which, in our experiments, reduces computations by 20% with equivalent or even superior accuracy on the ImageNet dataset, and significantly outperforms state-of-the-art approaches in terms of AP50 on the MS COCO object detection dataset. The CSPNet is easy to implement and general enough to cope with architectures based on ResNet, ResNeXt, and DenseNet. Source code is at https://github.com/WongKinYiu/CrossStagePartialNetworks.

PaperPDFCode

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

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

AlexeyAB/darknet officialmentioned in papermentioned on GitHubtfNOASSERTIONSyntology: not harvested report
WongKinYiu/CrossStagePartialNetworks officialmentioned in papermentioned on GitHubtfSyntology: not harvested report
6-dl/darknet_wpb mentioned on GitHubtfSyntology: not harvested report
Aarav-Patel/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
Abhijeet2396/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
Adam123wu/Edge-YOLO_algorithm mentioned on GitHubpytorchSyntology: not harvested report
AmeerHamza111/yolo-darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
CRIGIM/darknet mentioned on GitHubtfSyntology: not harvested report
Chamchis/AICCTV mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
Chris-On-PC/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
Code-Fight/darknet_study mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
Congpeiliang/yolo mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
Demonhesusheng/darknet_v2 mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
FarzamIHPC/YoloV3_mask mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
HSRW-EMRP-WS1920-Drone/object_detection_and_landing mentioned on GitHubtfSyntology: not harvested report
JCamERG/YOLOPersonal mentioned on GitHubtfSyntology: not harvested report
JEONGHA-SHIN/YOGO_Server_v2 mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
JWMON/yolo mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
Jorge-Mendes/darknet-google-colab mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
Keshavkant/YOLO_Toll mentioned on GitHubtfSyntology: not harvested report
Knight825/models-pytorch mentioned on GitHubpytorchApache-2.0Syntology: not harvested report
Pepitaw/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
PericlesPet/darknet-demo mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
PetersonSilvadeJesus/yolov3-data-images mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
Rahmanzia3/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
Rahmanzia3/darknet_original mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
Rahmanzia3/yolo mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
Ritik-Sharma38/YoloV3_training_Hands mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
SAIGANESH-Collab/DARKNET mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
SAIGANESH-Collab/YOLO-AQUA mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
Seioch/darknet_v4_deploy mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
Smiler-Jin/Yolov4 mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
Spain2394/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
SpanoChristian/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
Sujith2k/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
TusharMhaske28/Object-Detection mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
TusharMhaske28/Thermal-Object mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
TusharMhaske28/summer-project mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
WangMian-Maker/Flask mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
XuanjieXiao/vehicle-detection-by-using-camera mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
Yamahitsuji/darknet mentioned on GitHubtfSyntology: not harvested report
aaron-wang-de/ab mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
achen353/Taiwanese-Traffic-Object-Detection mentioned on GitHubtfSyntology: not harvested report
adam123wu/edge-yolo mentioned on GitHubpytorchSyntology: not harvested report
agelencs/darknet_original mentioned on GitHubtfSyntology: not harvested report
alexismailov2/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
amazingcodeLYL/Traffic_signs_detection_darket mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
antonoeschmidt/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
areberoto/darknet-thesis mentioned on GitHubtfSyntology: not harvested report
benjamintli/darknet-gun-detector mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
benjamintli/training-results-azure mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
bowensong0312/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
citbrains/darknet_yolov4 mentioned on GitHubtfSyntology: not harvested report
clasikman/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
dayfixer/darknet-comment mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
deeplearnproject/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
dipuchak95/dip_net mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
dorbodwolf/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
easyadin/Object-Detection-YOLOv4 mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
eric-erki/darknetAB mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
firja10/darknet-2 mentioned on GitHubtfSyntology: not harvested report
gomboskriszta/yolo mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
hirohiro23/Darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
holytrick/AlexeyAB_darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
iamsingh/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
iberganzo/darknet mentioned on GitHubtfSyntology: not harvested report
imshamsher/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
johnsebin97/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
k5iogura/darknetAB mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
kaustavsenkvs/Deep-Learning-Mini-Project mentioned on GitHubtfSyntology: not harvested report
kentttttt/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
khanh-moriaty/DarknetAlexey mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
kiruthick101/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
laclouis5/darknet2 mentioned on GitHubtfSyntology: not harvested report
lanmei1211/face mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
lichuanqi/YOLOV4_Darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
liuxiangchao369/VRS-training mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
lrf008/yolov4 mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
mazabou/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
meerkatai/Yolo-v4-Object-Detection-Classification mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
minhcong0891/wheat-detection mentioned on GitHubtfnot reachable when probed 2026-09-17 — repositories for recent papers often appear after camera-readySyntology: not harvested report
neung678/anti_bird mentioned on GitHubtfSyntology: not harvested report
neverleave0916/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
nghgphi/fire_detection mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
nmcasasr/darknet2 mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
nmcasasr/darknet_oldVersion mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
nurulxakmar/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
okboseong/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
otamajakusi/darknet-yolov4 mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
peter8301120/darknet_t mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
pixiedust18/darknet_mask_classification mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
pritish-mahajan/Apricitas mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
redwankarimsony/darknet-yolov4-working mentioned on GitHubtfSyntology: not harvested report
rwightman/pytorch-image-models mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to AdamP: Slowing Down the Slowdown for Momentum Optimizers on Scale-invariant Weights (arXiv:2006.08217; that paper's own run record: 4 ran (of which 1 constructed an object rather than computing a result; 2 with no instrument failure: 1 honoured, 0 violated, 1 with no contract checked; 2 where Syntology's instrument failed) · 0 unverified); nothing checks that they implement this paper's method. report
sadicLiu/darknet-yolov4 mentioned on GitHubtfSyntology: not harvested report
sepuhopar/ReportDD-darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
shanky1947/YOLOv3-Darknet-Custom-Object-Detection mentioned on GitHubtfSyntology: not harvested report
sryab2001/Yolo-CG mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
sryab2001/aquasaiclone mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
tanmayj000/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
tdurand/darknet-withtiling mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
texasinstruments/edgeai-yolov5 mentioned on GitHubpytorchGPL-3.0Syntology: not harvested report
texasinstruments/edgeai-yolox mentioned on GitHubpytorchNOASSERTIONSyntology: not harvested report
thodorostsao/4-leg-robot-neural-network mentioned on GitHubtfSyntology: not harvested report
tmbaoloc/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
tommyjtl/darknet-colab mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
tops666/volo-v4- mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
viskjha/Wrong-way-vehicle-detection mentioned on GitHubtfSyntology: not harvested report
wdlctc/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
witwickey/alexeyDarknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
wsmlby/darknetnano mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
xuantruongdang/module-yolo mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
xyuan-wyze/darknet mentioned on GitHubtfSyntology: not harvested report
yaoming0712/dalinmask mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
ycchiusieve/yolo3 mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
yukaryote/darknet mentioned on GitHubtfSyntology: not harvested report
zanmange/darknet mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
zephyrcho/darknetV4 mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
zhanghuiyao/yolov5_mindspore mentioned on GitHubmindsporeMITSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to YOLOv4: Optimal Speed and Accuracy of Object Detection (arXiv:2004.10934; that paper's own run record: 142 ran (of which 0 constructed an object rather than computing a result; 128 with no instrument failure: 8 honoured, 0 violated, 120 with no contract checked; 14 where Syntology's instrument failed) · 42 unverified); nothing checks that they implement this paper's method. report
Deci-AI/super-gradients pytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Deep Residual Learning for Image Recognition (arXiv:1512.03385; that paper's own run record: 254 ran (of which 108 constructed an object rather than computing a result; 163 with no instrument failure: 3 honoured, 1 violated, 159 with no contract checked; 91 where Syntology's instrument failed) · 123 unverified), MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications (arXiv:1704.04861; that paper's own run record: 53 ran (of which 28 constructed an object rather than computing a result; 44 with no instrument failure: 4 honoured, 0 violated, 40 with no contract checked; 9 where Syntology's instrument failed) · 30 unverified), RepVGG: Making VGG-style ConvNets Great Again (arXiv:2101.03697; that paper's own run record: 13 ran (of which 7 constructed an object rather than computing a result; 8 with no instrument failure: 0 honoured, 0 violated, 8 with no contract checked; 5 where Syntology's instrument failed) · 3 unverified) and 1 more (every one in this page's JSON twin); nothing checks that they implement this paper's method. report
PaddlePaddle/PaddleClas paddleApache-2.0Syntology: not harvested report
open-mmlab/mmclassification pytorchApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Reversible Vision Transformers (arXiv:2302.04869; that paper's own run record: 24 ran (of which 5 constructed an object rather than computing a result; 17 with no instrument failure: 0 honoured, 0 violated, 17 with no contract checked; 7 where Syntology's instrument failed) · 5 unverified); nothing checks that they implement this paper's method. report
gitlab.com/birder/birder pytorchSyntology: 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

AttributeImage ClassificationObject DetectionReal-Time Object Detection

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Image Classification ImageNet CSPResNeXt-50 (Mish+Aug) Number of params 20.5M #741 of 1060 Archive leaderboard report
Image Classification ImageNet CSPResNeXt-50 (Mish+Aug) Top 1 Accuracy 79.8% #741 of 1060 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

Introduced by this paper: CSPPeleeNet, CSPResNeXt, CSPResNeXt Block, Exact Fusion Model

1x1 ConvolutionAdaptive Feature PoolingAverage PoolingBatch NormalizationBottleneck Residual BlockBottom-up Path AugmentationCSPDarknet53CSPDenseNetCSPDenseNet-ElasticCSPPeleeNetCSPResNeXtCSPResNeXt BlockConcatenated Skip ConnectionConvolutionDense BlockDense ConnectionsDenseNet-ElasticDropoutElastic Dense BlockExact Fusion ModelFPNGlobal Average PoolingGrouped ConvolutionKaiming InitializationMax PoolingMaxoutPAFPNPANetPeleeNetPolynomial Rate DecayRPNReLUResNeXtResNeXt BlockResidual BlockResidual ConnectionRoIAlignSGD with MomentumSigmoid ActivationSoftmaxSqueeze-and-Excitation BlockStep DecayTanh ActivationTwo-Way Dense LayerWeight Decay

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