Papers › DenseNets Reloaded: Paradigm Shift Beyond ResNets and ViTs

DenseNets Reloaded: Paradigm Shift Beyond ResNets and ViTs

28 Mar 2024arXiv:2403.19588archive 2025-07-28

Donghyun Kim, Byeongho Heo, Dongyoon Han

This paper revives Densely Connected Convolutional Networks (DenseNets) and reveals the underrated effectiveness over predominant ResNet-style architectures. We believe DenseNets' potential was overlooked due to untouched training methods and traditional design elements not fully revealing their capabilities. Our pilot study shows dense connections through concatenation are strong, demonstrating that DenseNets can be revitalized to compete with modern architectures. We methodically refine suboptimal components - architectural adjustments, block redesign, and improved training recipes towards widening DenseNets and boosting memory efficiency while keeping concatenation shortcuts. Our models, employing simple architectural elements, ultimately surpass Swin Transformer, ConvNeXt, and DeiT-III - key architectures in the residual learning lineage. Furthermore, our models exhibit near state-of-the-art performance on ImageNet-1K, competing with the very recent models and downstream tasks, ADE20k semantic segmentation, and COCO object detection/instance segmentation. Finally, we provide empirical analyses that uncover the merits of the concatenation over additive shortcuts, steering a renewed preference towards DenseNet-style designs. Our code is available at https://github.com/naver-ai/rdnet.

PaperPDFCodeCode Syntology ran

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also 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

Syntology 3 code samples harvested from 1 repository linked to this paper: 3 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 3 honoured, 0 violated, 0 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code; here there are none.

By repository: official repository: 3 samples from 1 repository, 3 ran (3 with no instrument failure; 0 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

No sample is linked to this paper from 1 of its repositories that Syntology harvested for other papers:

No date is recorded for any of the 3 samples that ran: their levels were written before Syntology's graph kept the date.

huggingface/pytorch-image-models officialmentioned in papermentioned on GitHubpytorchApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Learning to Merge Tokens via Decoupled Embedding for Efficient Vision Transformers (arXiv:2412.10569; that paper's own run record: 28 ran (of which 1 constructed an object rather than computing a result; 12 with no instrument failure: 1 honoured, 0 violated, 11 with no contract checked; 16 where Syntology's instrument failed) · 6 unverified); nothing checks that they implement this paper's method. report
naver-ai/rdnet officialmentioned in papermentioned on GitHubpytorchApache-2.0Syntology: 3 samples, 3 ran (0 instrument-failed) · 0 unverified 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

3 samples harvested: 3 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 3 honoured, 0 violated, 0 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page has not changed since , the first build that kept a record date for it; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

3ran · honoured contract

Licence: 0 of the 3 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 naver-ai/rdnet. “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. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (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 are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. 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 six 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; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. 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.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity.

naver-ai/rdnet official repository 3 samples, 3 ran (0 instrument-failed) · 0 unverified
count_params in 80 papers benchmark.py ran · honoured contract Apache-2.0 (permissive) · 410bffb8a07b52e3 · report fixture: signature model:callable/0/? · returned int · level recorded, date not recorded
cuda_timestamp in 11 papers benchmark.py ran · honoured contract fingerprinted Apache-2.0 (permissive) · 50d653ee6a59955c · report fixture: signature sync:bool/0/bool, device:NoneType/0/? · returned float · level recorded, date not recorded
timestamp in 15 papers benchmark.py ran · honoured contract Apache-2.0 (permissive) · 475160c5a1a1225a · report fixture: signature sync:bool/0/bool · returned float · level recorded, date not recorded

Tasks

Fine-Grained Image ClassificationImage ClassificationInstance SegmentationObject DetectionSemantic Segmentationobject-detection

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Fine-Grained Image Classification Stanford Cars RDNet-S (224 res, IN-1K pretrained) Accuracy 94.2% #51 of 83 Archive leaderboard report
Fine-Grained Image Classification Stanford Cars RDNet-S (224 res, IN-1K pretrained) FLOPS 8.7G #51 of 83 Archive leaderboard report
Fine-Grained Image Classification Stanford Cars RDNet-S (224 res, IN-1K pretrained) PARAMS 50M #51 of 83 Archive leaderboard report
Fine-Grained Image Classification Stanford Cars RDNet-L (224 res, IN-1K pretrained) Accuracy 94.2% #52 of 83 Archive leaderboard report
Fine-Grained Image Classification Stanford Cars RDNet-L (224 res, IN-1K pretrained) FLOPS 34.7G #52 of 83 Archive leaderboard report
Fine-Grained Image Classification Stanford Cars RDNet-L (224 res, IN-1K pretrained) PARAMS 186M #52 of 83 Archive leaderboard report
Fine-Grained Image Classification Stanford Cars RDNet-B (224 res, IN-1K pretrained) Accuracy 94.1% #53 of 83 Archive leaderboard report
Fine-Grained Image Classification Stanford Cars RDNet-B (224 res, IN-1K pretrained) FLOPS 15.4G #53 of 83 Archive leaderboard report
Fine-Grained Image Classification Stanford Cars RDNet-B (224 res, IN-1K pretrained) PARAMS 87M #53 of 83 Archive leaderboard report
Fine-Grained Image Classification Stanford Cars RDNet-T (224 res, IN-1K pretrained) Accuracy 93.9% #59 of 83 Archive leaderboard report
Fine-Grained Image Classification Stanford Cars RDNet-T (224 res, IN-1K pretrained) FLOPS 5.0G #59 of 83 Archive leaderboard report
Fine-Grained Image Classification Stanford Cars RDNet-T (224 res, IN-1K pretrained) PARAMS 24M #59 of 83 Archive leaderboard report
Image Classification CIFAR-10 RDNet-L (224 res, IN-1K pretrained) Percentage correct 99.31 #8 of 265 Archive leaderboard report
Image Classification CIFAR-10 RDNet-B (224 res, IN-1K pretrained) Percentage correct 99.31 #9 of 265 Archive leaderboard report
Image Classification CIFAR-10 RDNet-T (224 res, IN-1K pretrained) Percentage correct 98.88 #28 of 265 Archive leaderboard report
Image Classification ImageNet RDNet-L (384 res) GFLOPs 34.7 #197 of 1060 Archive leaderboard report
Image Classification ImageNet RDNet-L (384 res) Number of params 186M #197 of 1060 Archive leaderboard report
Image Classification ImageNet RDNet-L (384 res) Top 1 Accuracy 85.8% #197 of 1060 Archive leaderboard report
Image Classification ImageNet RDNet-L GFLOPs 34.7 #292 of 1060 Archive leaderboard report
Image Classification ImageNet RDNet-L Number of params 186M #292 of 1060 Archive leaderboard report
Image Classification ImageNet RDNet-L Top 1 Accuracy 84.8% #292 of 1060 Archive leaderboard report
Image Classification ImageNet RDNet-B GFLOPs 15.4 #320 of 1060 Archive leaderboard report
Image Classification ImageNet RDNet-B Number of params 87M #320 of 1060 Archive leaderboard report
Image Classification ImageNet RDNet-B Top 1 Accuracy 84.4% #320 of 1060 Archive leaderboard report
Image Classification ImageNet RDNet-S GFLOPs 8.7 #395 of 1060 Archive leaderboard report
Image Classification ImageNet RDNet-S Number of params 50M #395 of 1060 Archive leaderboard report
Image Classification ImageNet RDNet-S Top 1 Accuracy 83.7% #395 of 1060 Archive leaderboard report
Image Classification ImageNet RDNet-T GFLOPs 5.0 #496 of 1060 Archive leaderboard report
Image Classification ImageNet RDNet-T Number of params 24M #496 of 1060 Archive leaderboard report
Image Classification ImageNet RDNet-T Top 1 Accuracy 82.8% #496 of 1060 Archive leaderboard report
Image Classification iNaturalist 2018 RDNet-L (224 res, IN-1K pretrained) Number of params 186M #12 of 60 Archive leaderboard report
Image Classification iNaturalist 2018 RDNet-L (224 res, IN-1K pretrained) Top-1 Accuracy 81.8% #12 of 60 Archive leaderboard report
Image Classification iNaturalist 2018 RDNet-B (224 res, IN-1K pretrained) Number of params 87M #16 of 60 Archive leaderboard report
Image Classification iNaturalist 2018 RDNet-B (224 res, IN-1K pretrained) Top-1 Accuracy 80.5 #16 of 60 Archive leaderboard report
Image Classification iNaturalist 2018 RDNet-S (224 res, IN-1K pretrained) Number of params 50M #20 of 60 Archive leaderboard report
Image Classification iNaturalist 2018 RDNet-S (224 res, IN-1K pretrained) Top-1 Accuracy 79.1 #20 of 60 Archive leaderboard report
Image Classification iNaturalist 2018 RDNet-T (224 res, IN-1K pretrained) Number of params 24M #24 of 60 Archive leaderboard report
Image Classification iNaturalist 2018 RDNet-T (224 res, IN-1K pretrained) Top-1 Accuracy 77.0 #24 of 60 Archive leaderboard report
Image Classification iNaturalist 2019 RDNet-L (224 res, IN-1K pretrained) Number of params 186M #5 of 22 Archive leaderboard report
Image Classification iNaturalist 2019 RDNet-L (224 res, IN-1K pretrained) Top-1 Accuracy 83.7 #5 of 22 Archive leaderboard report
Image Classification iNaturalist 2019 RDNet-B (224 res, IN-1K pretrained) Number of params 87M #6 of 22 Archive leaderboard report
Image Classification iNaturalist 2019 RDNet-B (224 res, IN-1K pretrained) Top-1 Accuracy 83.5 #6 of 22 Archive leaderboard report
Image Classification iNaturalist 2019 RDNet-S (224 res, IN-1K pretrained) Number of params 50M #7 of 22 Archive leaderboard report
Image Classification iNaturalist 2019 RDNet-S (224 res, IN-1K pretrained) Top-1 Accuracy 82.9 #7 of 22 Archive leaderboard report
Image Classification iNaturalist 2019 RDNet-T (224 res, IN-1K pretrained) Number of params 24M #11 of 22 Archive leaderboard report
Image Classification iNaturalist 2019 RDNet-T (224 res, IN-1K pretrained) Top-1 Accuracy 81.2 #11 of 22 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: RDNet

Absolute Position EncodingsAdamAttentionBPEConvNeXtDense ConnectionsDropoutLabel SmoothingLayer NormalizationLinear LayerMulti-Head AttentionPosition-Wise Feed-Forward LayerRDNetResidual ConnectionSoftmaxStochastic DepthSwin TransformerTransformer

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