Papers › Simple Training Strategies and Model Scaling for Object Detection

Simple Training Strategies and Model Scaling for Object Detection

30 Jun 2021arXiv:2107.00057archive 2025-07-28

Xianzhi Du, Barret Zoph, Wei-Chih Hung, Tsung-Yi Lin

The speed-accuracy Pareto curve of object detection systems have advanced through a combination of better model architectures, training and inference methods. In this paper, we methodically evaluate a variety of these techniques to understand where most of the improvements in modern detection systems come from. We benchmark these improvements on the vanilla ResNet-FPN backbone with RetinaNet and RCNN detectors. The vanilla detectors are improved by 7.7% in accuracy while being 30% faster in speed. We further provide simple scaling strategies to generate family of models that form two Pareto curves, named RetinaNet-RS and Cascade RCNN-RS. These simple rescaled detectors explore the speed-accuracy trade-off between the one-stage RetinaNet detectors and two-stage RCNN detectors. Our largest Cascade RCNN-RS models achieve 52.9% AP with a ResNet152-FPN backbone and 53.6% with a SpineNet143L backbone. Finally, we show the ResNet architecture, with three minor architectural changes, outperforms EfficientNet as the backbone for object detection and instance segmentation systems.

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

tensorflow/tpu officialmentioned in papertfApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Towards Understanding Why Label Smoothing Degrades Selective Classification and How to Fix It (arXiv:2403.14715; that paper's own run record: 9 ran (of which 2 constructed an object rather than computing a result; 6 with no instrument failure: 1 honoured, 0 violated, 5 with no contract checked; 3 where Syntology's instrument failed) · 1 unverified); nothing checks that they implement this paper's method. 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

Instance SegmentationObjectObject DetectionSemantic Segmentationobject-detection

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Object Detection COCO minival Cascade RCNN-RS (SpineNet-143L, single scale) APL 70.6 #59 of 220 Archive leaderboard report
Object Detection COCO minival Cascade RCNN-RS (SpineNet-143L, single scale) APM 56.7 #59 of 220 Archive leaderboard report
Object Detection COCO minival Cascade RCNN-RS (SpineNet-143L, single scale) APS 34.5 #59 of 220 Archive leaderboard report
Object Detection COCO minival Cascade RCNN-RS (SpineNet-143L, single scale) box AP 53.6 #59 of 220 Archive leaderboard report
Object Detection COCO minival Cascade RCNN-RS (ResNet-200, single scale) APL 70.3 #62 of 220 Archive leaderboard report
Object Detection COCO minival Cascade RCNN-RS (ResNet-200, single scale) APM 56.2 #62 of 220 Archive leaderboard report
Object Detection COCO minival Cascade RCNN-RS (ResNet-200, single scale) APS 33.9 #62 of 220 Archive leaderboard report
Object Detection COCO minival Cascade RCNN-RS (ResNet-200, single scale) box AP 53.1 #62 of 220 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 NormalizationBottleneck Residual BlockConvolutionDense ConnectionsDepthwise ConvolutionDepthwise Separable ConvolutionDropoutFPNFocal LossGlobal Average PoolingInverted Residual BlockKaiming InitializationMax PoolingPointwise ConvolutionRMSPropReLUResNet-DResidual BlockResidual ConnectionRetinaNetRetinaNet-RSSiLUSigmoid ActivationSqueeze-and-Excitation BlockXavier Initialization

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