Papers › Scaled-YOLOv4: Scaling Cross Stage Partial Network

Scaled-YOLOv4: Scaling Cross Stage Partial Network

16 Nov 2020CVPR 2021 1arXiv:2011.08036archive 2025-07-28

Chien-Yao Wang, Alexey Bochkovskiy, Hong-Yuan Mark Liao

We show that the YOLOv4 object detection neural network based on the CSP approach, scales both up and down and is applicable to small and large networks while maintaining optimal speed and accuracy. We propose a network scaling approach that modifies not only the depth, width, resolution, but also structure of the network. YOLOv4-large model achieves state-of-the-art results: 55.5% AP (73.4% AP50) for the MS COCO dataset at a speed of ~16 FPS on Tesla V100, while with the test time augmentation, YOLOv4-large achieves 56.0% AP (73.3 AP50). To the best of our knowledge, this is currently the highest accuracy on the COCO dataset among any published work. The YOLOv4-tiny model achieves 22.0% AP (42.0% AP50) at a speed of 443 FPS on RTX 2080Ti, while by using TensorRT, batch size = 4 and FP16-precision the YOLOv4-tiny achieves 1774 FPS.

PaperPDFConference PDFCodeCode Syntology ran

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

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="2011.08036")

Code

Syntology Ran 1 of 1 code samples harvested from 0 repositories linked to this paper; 0 have no recorded run. Of those that ran: 1 ran · our draft was wrong.

By repository: 1 identical to code first harvested elsewhere. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

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

WongKinYiu/ScaledYOLOv4 officialmentioned on GitHubpytorch report
6-dl/darknet_wpb mentioned on GitHubtf report
Adam123wu/Edge-YOLO_algorithm mentioned on GitHubpytorch report
AlexeyAB/darknet mentioned on GitHubtfNOASSERTION report
CRIGIM/darknet mentioned on GitHubtf report
CSSFA123/yolov4-Road mentioned on GitHubpytorch report
Eric3911/ScaledYOLOv4 mentioned on GitHubpytorch report
JCamERG/YOLOPersonal mentioned on GitHubtf report
RangiLyu/nanodet mentioned on GitHubpytorch report
Yamahitsuji/darknet mentioned on GitHubtf report
adam123wu/edge-yolo mentioned on GitHubpytorch report
adamlakatos0407/ScaledYOLOv4 mentioned on GitHubpytorch report
agelencs/darknet_original mentioned on GitHubtf report
areberoto/darknet-thesis mentioned on GitHubtf report
citbrains/darknet_yolov4 mentioned on GitHubtf report
firja10/darknet-2 mentioned on GitHubtf report
gaurav67890/ScaledYOLO mentioned on GitHubpytorch report
husty530/Yamashita-darknet mentioned on GitHub report
iberganzo/darknet mentioned on GitHubtf report
ikuokuo/start-scaled-yolov4 mentioned on GitHubpytorch report
joker311200/yolov4-csp mentioned on GitHubpytorch report
kyip375/ScaledYOLOv4 mentioned on GitHub report
laclouis5/darknet2 mentioned on GitHubtf report
leggedrobotics/darknet mentioned on GitHubpytorch report
muraliadithya/vdp mentioned on GitHubpytorch report
neung678/anti_bird mentioned on GitHubtf report
rjlallana/YOLO mentioned on GitHubpytorch report
xolbynz/scaled_yolo mentioned on GitHubpytorch report
xyuan-wyze/darknet mentioned on GitHubtf report
yousefis/darknet-ocr mentioned on GitHubpytorch report
yukaryote/darknet mentioned on GitHubtf 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.

Code Syntology ran Syntology

1 sample harvested; 1 ran; 0 honoured the contract we drafted; 0 have no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

1ran · our draft was wrong

Licence: 1 of the 1 sample is pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Some samples are identical code Syntology first harvested from another repository; for those, this paper's copy is not located and its licence is not recorded. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples from a repository marked official are listed first. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of four reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

load_classes identical code first harvested elsewhere ran · our draft was wrong licence of this copy not recorded · fa99a26b6ba8f205 · report

Tasks

2D Object DetectionObject DetectionReal-Time Object Detection

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Object Detection COCO minival YOLOv4-P7 CSP-P7 (single-scale, 16 fps) AP50 73.3 #49 of 220 Archive leaderboard report
Object Detection COCO minival YOLOv4-P7 CSP-P7 (single-scale, 16 fps) AP75 60.7 #49 of 220 Archive leaderboard report
Object Detection COCO minival YOLOv4-P7 CSP-P7 (single-scale, 16 fps) APL 67.4 #49 of 220 Archive leaderboard report
Object Detection COCO minival YOLOv4-P7 CSP-P7 (single-scale, 16 fps) APM 59.5 #49 of 220 Archive leaderboard report
Object Detection COCO minival YOLOv4-P7 CSP-P7 (single-scale, 16 fps) APS 38.1 #49 of 220 Archive leaderboard report
Object Detection COCO minival YOLOv4-P7 CSP-P7 (single-scale, 16 fps) box AP 55.4 #49 of 220 Archive leaderboard report
Object Detection COCO test-dev YOLOv4-P7 with TTA AP50 73.2 #47 of 225 Archive leaderboard report
Object Detection COCO test-dev YOLOv4-P7 with TTA AP75 61.2 #47 of 225 Archive leaderboard report
Object Detection COCO test-dev YOLOv4-P7 with TTA box mAP 55.8 #47 of 225 Archive leaderboard report
Object Detection COCO test-dev YOLOv4-P6 with TTA AP50 72.6 #50 of 225 Archive leaderboard report
Object Detection COCO test-dev YOLOv4-P6 with TTA AP75 60.2 #50 of 225 Archive leaderboard report
Object Detection COCO test-dev YOLOv4-P6 with TTA box mAP 54.9 #50 of 225 Archive leaderboard report
Object Detection COCO test-dev YOLOv4-P6 CSP-P6 (single-scale, 32 fps) AP50 72.3 #55 of 225 Archive leaderboard report
Object Detection COCO test-dev YOLOv4-P6 CSP-P6 (single-scale, 32 fps) AP75 59.5 #55 of 225 Archive leaderboard report
Object Detection COCO test-dev YOLOv4-P6 CSP-P6 (single-scale, 32 fps) APL 65.5 #55 of 225 Archive leaderboard report
Object Detection COCO test-dev YOLOv4-P6 CSP-P6 (single-scale, 32 fps) APM 58.2 #55 of 225 Archive leaderboard report
Object Detection COCO test-dev YOLOv4-P6 CSP-P6 (single-scale, 32 fps) APS 36.6 #55 of 225 Archive leaderboard report
Object Detection COCO test-dev YOLOv4-P6 CSP-P6 (single-scale, 32 fps) box mAP 54.3 #55 of 225 Archive leaderboard report
Object Detection COCO test-dev YOLOv4-P5 with TTA AP50 70.3 #70 of 225 Archive leaderboard report
Object Detection COCO test-dev YOLOv4-P5 with TTA AP75 58 #70 of 225 Archive leaderboard report
Object Detection COCO test-dev YOLOv4-P5 with TTA box mAP 52.5 #70 of 225 Archive leaderboard report
Object Detection COCO test-dev YOLOv4 (CD53) AP50 64.1 #138 of 225 Archive leaderboard report
Object Detection COCO test-dev YOLOv4 (CD53) AP75 49.5 #138 of 225 Archive leaderboard report
Object Detection COCO test-dev YOLOv4 (CD53) APL 56.7 #138 of 225 Archive leaderboard report
Object Detection COCO test-dev YOLOv4 (CD53) APM 49 #138 of 225 Archive leaderboard report
Object Detection COCO test-dev YOLOv4 (CD53) APS 27 #138 of 225 Archive leaderboard report
Object Detection COCO test-dev YOLOv4 (CD53) box mAP 45.5 #138 of 225 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 ConvolutionAverage PoolingBatch NormalizationBottom-up Path AugmentationCSPDarknet53ConvolutionCosine AnnealingCutMixDropBlockFPNGlobal Average PoolingGrid SensitiveLabel SmoothingLogistic RegressionMax PoolingPAFPNReLUResidual ConnectionSigmoid ActivationSoftmaxSpatial Pyramid PoolingTanh ActivationYOLOv3YOLOv4k-Means Clustering

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