Papers › Dynamic R-CNN: Towards High Quality Object Detection via Dynamic Training

Dynamic R-CNN: Towards High Quality Object Detection via Dynamic Training

13 Apr 2020ECCV 2020 8arXiv:2004.06002archive 2025-07-28

Hongkai Zhang, Hong Chang, Bingpeng Ma, Naiyan Wang, Xilin Chen

Although two-stage object detectors have continuously advanced the state-of-the-art performance in recent years, the training process itself is far from crystal. In this work, we first point out the inconsistency problem between the fixed network settings and the dynamic training procedure, which greatly affects the performance. For example, the fixed label assignment strategy and regression loss function cannot fit the distribution change of proposals and thus are harmful to training high quality detectors. Consequently, we propose Dynamic R-CNN to adjust the label assignment criteria (IoU threshold) and the shape of regression loss function (parameters of SmoothL1 Loss) automatically based on the statistics of proposals during training. This dynamic design makes better use of the training samples and pushes the detector to fit more high quality samples. Specifically, our method improves upon ResNet-50-FPN baseline with 1.9% AP and 5.5% AP₉₀ on the MS COCO dataset with no extra overhead. Codes and models are available at https://github.com/hkzhang95/DynamicRCNN.

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="2004.06002")

Code

Syntology Ran 2 of 18 code samples harvested from 2 repositories linked to this paper; 16 have no recorded run. Of those that ran: 1 ran · violated contract; 1 ran with no contract checked.

By repository: official repository: 4 samples from 1 repository, 2 ran; community (archive-listed): 14 samples from 1 repository, 0 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

hkzhang95/DynamicRCNN officialmentioned in papermentioned on GitHubpytorchMIT report
TangShengqin/DynamicRCNN-mmdet mentioned on GitHubpytorchApache-2.0 report
open-mmlab/mmdetection pytorchApache-2.0 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

18 samples harvested; 2 ran; 0 honoured the contract we drafted; 16 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 · violated contract
1ran
16unverified

Licence: 0 of the 18 samples are 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.

Harvested from 2 repositories linked to this paper, official or community; each sample names its own and says which. “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.

has_valid_annotation hkzhang95/DynamicRCNN/dynamic_rcnn/datasets/coco.py official repository ran · violated contract MIT (permissive) · 2531c930ac17dddf · report
smooth_l1_loss hkzhang95/DynamicRCNN/dynamic_rcnn/det_opr/loss.py official repository ran MIT (permissive) · e261fa29066b37e5 · report
load_resnet_c2_format hkzhang95/DynamicRCNN/dynamic_rcnn/basemodels/c2_model_loading.py official repository unverified MIT (permissive) · b99b85e38568b360 · report
sigmoid_focal_loss_cpu hkzhang95/DynamicRCNN/dynamic_rcnn/det_opr/loss.py official repository unverified MIT (permissive) · 426fd02a2beaf826 · report
anchor_target TangShengqin/DynamicRCNN-mmdet/mmdet/core/anchor/anchor_target.py community (archive-listed) unverified Apache-2.0 (permissive) · 6a7ed994a87db611 · report
anchor_target_single TangShengqin/DynamicRCNN-mmdet/mmdet/core/anchor/anchor_target.py community (archive-listed) unverified Apache-2.0 (permissive) · 1dd00ccb42e850dc · report
bbox_target TangShengqin/DynamicRCNN-mmdet/mmdet/core/bbox/bbox_target.py community (archive-listed) unverified Apache-2.0 (permissive) · 34a7030ddea70bdd · report
bbox_target_single TangShengqin/DynamicRCNN-mmdet/mmdet/core/bbox/bbox_target.py community (archive-listed) unverified Apache-2.0 (permissive) · a66c7009b985da0a · report
calc_region TangShengqin/DynamicRCNN-mmdet/mmdet/core/anchor/guided_anchor_target.py community (archive-listed) unverified Apache-2.0 (permissive) · 736c7ce8a0faa02a · report
ensure_rng TangShengqin/DynamicRCNN-mmdet/mmdet/core/bbox/demodata.py community (archive-listed) unverified Apache-2.0 (permissive) · 8f43d2c26227a141 · report
expand_target TangShengqin/DynamicRCNN-mmdet/mmdet/core/bbox/bbox_target.py community (archive-listed) unverified Apache-2.0 (permissive) · a57ba57f6dde621b · report
ga_loc_target TangShengqin/DynamicRCNN-mmdet/mmdet/core/anchor/guided_anchor_target.py community (archive-listed) unverified Apache-2.0 (permissive) · 2440313353f0c835 · report
ga_shape_target TangShengqin/DynamicRCNN-mmdet/mmdet/core/anchor/guided_anchor_target.py community (archive-listed) unverified Apache-2.0 (permissive) · e3463fa8772a18a9 · report
images_to_levels TangShengqin/DynamicRCNN-mmdet/mmdet/core/anchor/anchor_target.py community (archive-listed) unverified Apache-2.0 (permissive) · 8fa249eba3a0ce57 · report
images_to_levels TangShengqin/DynamicRCNN-mmdet/mmdet/core/anchor/point_target.py community (archive-listed) unverified Apache-2.0 (permissive) · 99bbbfc95abdc86a · report
point_target TangShengqin/DynamicRCNN-mmdet/mmdet/core/anchor/point_target.py community (archive-listed) unverified Apache-2.0 (permissive) · a08ee98e88ad24e7 · report
point_target_single TangShengqin/DynamicRCNN-mmdet/mmdet/core/anchor/point_target.py community (archive-listed) unverified Apache-2.0 (permissive) · b660461f38b4382b · report
random_boxes TangShengqin/DynamicRCNN-mmdet/mmdet/core/bbox/demodata.py community (archive-listed) unverified Apache-2.0 (permissive) · 3dc30523833b925a · report

Tasks

Object DetectionVocal Bursts Intensity Predictionobject-detectionregression

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Object Detection COCO test-dev Dynamic R-CNN (ResNet-101-DCN, multi-scale) AP50 68.3 #93 of 225 Archive leaderboard report
Object Detection COCO test-dev Dynamic R-CNN (ResNet-101-DCN, multi-scale) AP75 55.6 #93 of 225 Archive leaderboard report
Object Detection COCO test-dev Dynamic R-CNN (ResNet-101-DCN, multi-scale) APL 61.2 #93 of 225 Archive leaderboard report
Object Detection COCO test-dev Dynamic R-CNN (ResNet-101-DCN, multi-scale) APM 53.0 #93 of 225 Archive leaderboard report
Object Detection COCO test-dev Dynamic R-CNN (ResNet-101-DCN, multi-scale) APS 32.8 #93 of 225 Archive leaderboard report
Object Detection COCO test-dev Dynamic R-CNN (ResNet-101-DCN, multi-scale) box mAP 50.1 #93 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

Introduced by this paper: Dynamic R-CNN

1x1 ConvolutionAverage PoolingBatch NormalizationBottleneck Residual BlockConvolutionDeformable ConvolutionDynamic R-CNNDynamic SmoothL1 LossFPNFaster R-CNNGlobal Average PoolingKaiming InitializationMax PoolingNon Maximum SuppressionRPNReLUResidual BlockResidual ConnectionRoIPoolSoft-NMSSoftmax

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