Methods › General › Normalization › Group Normalization
Group Normalization
Introduced by Yuxin Wu et al. in Group Normalization
archive 2025-07-28 Description, source and code snippet are the archive's method entry.
Group Normalization is a normalization layer that divides channels into groups and normalizes the features within each group. GN does not exploit the batch dimension, and its computation is independent of batch sizes. In the case where the group size is 1, it is equivalent to Instance Normalization.
As motivation for the method, many classical features like SIFT and HOG had group-wise features and involved group-wise normalization. For example, a HOG vector is the outcome of several spatial cells where each cell is represented by a normalized orientation histogram.
Formally, Group Normalization is defined as:
μᵢ = 1/m∑_(k∈𝒮ᵢ)xₖ
σ²ᵢ = 1/m∑_(k∈𝒮ᵢ)(xₖ-μᵢ)²
x̂ᵢ = (xᵢ - μᵢ)/(√(σ²ᵢ+ϵ))
Here x is the feature computed by a layer, and i is an index. Formally, a Group Norm layer computes μ and σ in a set 𝒮ᵢ defined as: 𝒮ᵢ ={k |k_N = i_N ,⌊k_C/(C/G)⌋= ⌊I_C/(C/G)⌋}.
Here G is the number of groups, which is a pre-defined hyper-parameter (G = 32 by default). C/G is the number of channels per group. ⌊ is the floor operation, and the final term means that the indexes i and k are in the same group of channels, assuming each group of channels are stored in a sequential order along the C axis.
In Syntology repository known, no verified implementation (declared_code; repositories: 5 listed of 23 in the graph, strongest link: official). This answer is about arXiv:1803.08494, the paper the archive credits with this method.
Code Syntology holds under other names, as the lookup lists them (code_nearby); not a complete list, and none of them is this method.
- SELECTIVE TASK GROUP UPDATES · shares_name_tokens (shares name tokens with this name; a different method) · level V2 · retrieve with
get_reference_implementation(name='SELECTIVE TASK GROUP UPDATES') - new group invariant learning method · shares_name_tokens (shares name tokens with this name; a different method) · level V3 · retrieve with
get_reference_implementation(name='new group invariant learning method') - Group Preference Optimization (GPO) · shares_name_tokens (shares name tokens with this name; a different method) · level V3 · retrieve with
get_reference_implementation(name='Group Preference Optimization (GPO)') - Group-robust Sample Reweighting (GSR) · shares_name_tokens (shares name tokens with this name; a different method) · level V2 · retrieve with
get_reference_implementation(name='Group-robust Sample Reweighting (GSR)') - Frequency Adaptive Normalization (FAN) · shares_name_tokens (shares name tokens with this name; a different method) · level V3 · retrieve with
get_reference_implementation(name='Frequency Adaptive Normalization (FAN)') - Adversarial Batch Normalization (AdvBN) · shares_name_tokens (shares name tokens with this name; a different method) · level V3 · retrieve with
get_reference_implementation(name='Adversarial Batch Normalization (AdvBN)')
Origin paper of this answer, arXiv:1803.08494 (its page on this site · its page on Syntology): Syntology's run record for its harvested code, 7 ran (of which 2 constructed an object rather than computing a result; 5 with no instrument failure: 0 honoured, 0 violated, 5 with no contract checked; 2 where Syntology's instrument failed) · 8 unverified (of 15 harvested samples) · 5 pointer-only (licence). The instrument figure counts failures of Syntology's instrument, not of the code. Ran means executed on a synthesized input, not that the code is correct.
Syntology's have check for this name (free, it answers coverage only), read 2026-10-05; ask it again live. A level is the served verification level of that one answer; nothing here is summed across methods.
What each level means, quoted from Syntology's server card: V0 “imports”; V1 “executes against a synthesized fixture and returns a value of the recorded type”; V2 “two implementations GENERATED from the paper's signature and spec agree numerically (1e-6) on fixture inputs” “neither of the two ever read the authors' own code, so V2 measures how unambiguously the paper specifies the method, not fidelity to the authors' implementation (measured 2026-09-11, commit 75eaa77: cross_checked_by is a second LLM on 2,628 of 2,831 served samples)”; V3 “paper-derived property tests pass on both of those same two generations.” The card adds: “No rung on this ladder compares the served code against the authors' implementation, so treat every level as evidence about runnability and internal agreement, not about matching the paper's results.” A level quoted in two parts is one definition the card writes as two clauses joined by a double hyphen, each clause quoted on its own (how these levels are quoted).
Papers archive 2025-07-28
archive papers tagged: 55 · with a code link: 31 · where Syntology ran a sample: 16 (13 with a run with no instrument failure, 3 where every run was a failure of Syntology's instrument) Syntology
Show: all tagged papersonly where code ran (16 of 55 tagged: 13 with a run with no instrument failure, 3 where every run was a failure of Syntology's instrument)
Page 1 of 1: papers 1 to 55 of 55, newest first by the archive's date (ties by slug), in archive order.
Repository counts are the archive's code-links table. A Syntology line states what Syntology ran from that paper's harvested code, as “N ran (of which C constructed an object rather than computing a result; K with no instrument failure: H honoured, V violated, P with no contract checked; I where Syntology's instrument failed) · U unverified”; the instrument figure counts failures of Syntology's instrument, not of the code. It is per sample and not a correctness claim. When the archive marks a repository official for the paper, the line starts with that repository's state (the archive's flag, not a verdict on who wrote the code; “community repositories only” when every sample that ran came from a community repository, “official: no sample here; runs from other or unrecorded repositories” when some came from a repository the paper names or has in its text, or from none recorded); hover it for the repositories the samples that ran came from.
-
Dynamic Group Normalization: Spatio-Temporal Adaptation to Evolving Data Statistics 1 Jan 2025 · 0 repositories
-
Rethinking Normalization Strategies and Convolutional Kernels for Multimodal Image Fusion 15 Nov 2024 · 0 repositories · arXiv:2411.10036
-
Unsupervised Adaptive Normalization 7 Sep 2024 · 1 repository · arXiv:2409.04757
-
Exploring the Efficacy of Group-Normalization in Deep Learning Models for Alzheimer's Disease Classification 1 Apr 2024 · 0 repositories · arXiv:2404.00946
-
Training-Free Pretrained Model Merging 4 Mar 2024 · 1 repository · arXiv:2403.01753Syntology official (archive's flag): 3 ran · 3 ran (of which 1 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 2 where Syntology's instrument failed) · 1 unverified (of 4 harvested samples) · 4 pointer-only (licence)
-
ELA: Efficient Local Attention for Deep Convolutional Neural Networks 2 Mar 2024 · 0 repositories · arXiv:2403.01123
-
On Sensitivity and Robustness of Normalization Schemes to Input Distribution Shifts in Automatic MR Image Diagnosis 23 Jun 2023 · 1 repository · arXiv:2306.13276Syntology official (archive's flag): 1 ran · 1 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified (of 1 harvested sample) · 1 pointer-only (licence)
-
Adaptive Sparse Convolutional Networks with Global Context Enhancement for Faster Object Detection on Drone Images 25 Mar 2023 · 1 repository · arXiv:2303.14488Syntology official (archive's flag): 4 ran · 4 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 2 where Syntology's instrument failed) · 2 unverified (of 6 harvested samples) · 4 pointer-only (licence)
-
Making Batch Normalization Great in Federated Deep Learning 12 Mar 2023 · 0 repositories · arXiv:2303.06530
-
On the Ideal Number of Groups for Isometric Gradient Propagation 7 Feb 2023 · 0 repositories · arXiv:2302.03193
-
MGTUNet: An new UNet for colon nuclei instance segmentation and quantification 20 Oct 2022 · 0 repositories · arXiv:2210.10981
-
Kernel Normalized Convolutional Networks for Privacy-Preserving Machine Learning 30 Sep 2022 · 0 repositories · arXiv:2210.00053
-
Training a universal instance segmentation network for live cell images of various cell types and imaging modalities 28 Jul 2022 · 1 repository · arXiv:2207.14347
-
Understanding and Improving Group Normalization 5 Jul 2022 · 1 repository · arXiv:2207.01972
-
Domain Adaptation and Active Learning for Fine-Grained Recognition in the Field of Biodiversity 22 Oct 2021 · 0 repositories · arXiv:2110.11778
-
Exploring Heterogeneous Characteristics of Layers in ASR Models for More Efficient Training 8 Oct 2021 · 0 repositories · arXiv:2110.04267
-
Scalable deeper graph neural networks for high-performance materials property prediction 25 Sep 2021 · 1 repository · arXiv:2109.12283
-
NanoBatch Privacy: Enabling fast Differentially Private learning on the IPU 24 Sep 2021 · 0 repositories · arXiv:2109.12191
-
Benchmarking the Robustness of Instance Segmentation Models 2 Sep 2021 · 0 repositories · arXiv:2109.01123
-
Effect of Pre-Training Scale on Intra- and Inter-Domain Full and Few-Shot Transfer Learning for Natural and Medical X-Ray Chest Images 31 May 2021 · 1 repository · arXiv:2106.00116
-
GENESIS-V2: Inferring Unordered Object Representations without Iterative Refinement 20 Apr 2021 · 2 repositories · arXiv:2104.09958
-
Adaptive Feature Fusion Network for Gaze Tracking in Mobile Tablets 20 Mar 2021 · 0 repositories · arXiv:2103.11119
-
SpecTr: Spectral Transformer for Hyperspectral Pathology Image Segmentation 5 Mar 2021 · 1 repository · arXiv:2103.03604
-
Batch Group Normalization 4 Dec 2020 · 0 repositories · arXiv:2012.02782
-
NLCA-Net v2 for Stereo Matching in ECCV'20 Robust Vision Challenge 1 Nov 2020 · 0 repositories
-
High resolution weakly supervised localization architectures for medical images 22 Oct 2020 · 1 repository · arXiv:2010.11475
-
BYOL works even without batch statistics 20 Oct 2020 · 3 repositories · arXiv:2010.10241Syntology 5 ran (of which 0 constructed an object rather than computing a result; 4 with no instrument failure: 1 honoured, 3 violated, 0 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified (of 5 harvested samples) · 3 pointer-only (licence)
-
Group Whitening: Balancing Learning Efficiency and Representational Capacity 28 Sep 2020 · 1 repository · arXiv:2009.13333Syntology 5 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 3 where Syntology's instrument failed) · 0 unverified (of 5 harvested samples) · 1 pointer-only (licence)
-
Improving Semi-supervised Federated Learning by Reducing the Gradient Diversity of Models 26 Aug 2020 · 1 repository · arXiv:2008.11364Syntology official (archive's flag): 6 ran · 6 ran (of which 0 constructed an object rather than computing a result; 4 with no instrument failure: 2 honoured, 0 violated, 2 with no contract checked; 2 where Syntology's instrument failed) · 5 unverified (of 11 harvested samples) · 3 pointer-only (licence)
-
New Interpretations of Normalization Methods in Deep Learning 16 Jun 2020 · 0 repositories · arXiv:2006.09104
-
Towards Deeper Graph Neural Networks with Differentiable Group Normalization 12 Jun 2020 · 1 repository · arXiv:2006.06972
-
Effective Data Fusion with Generalized Vegetation Index: Evidence from Land Cover Segmentation in Agriculture 7 May 2020 · 0 repositories · arXiv:2005.03743
-
Evolving Normalization-Activation Layers 6 Apr 2020 · 8 repositories · arXiv:2004.02967Syntology 3 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 3 where Syntology's instrument failed) · 2 unverified (of 5 harvested samples) · 3 pointer-only (licence)
-
Pipelined Backpropagation at Scale: Training Large Models without Batches 25 Mar 2020 · 0 repositories · arXiv:2003.11666
-
Extended Batch Normalization 12 Mar 2020 · 0 repositories · arXiv:2003.05569
-
Self-Supervised Visual Learning by Variable Playback Speeds Prediction of a Video 5 Mar 2020 · 1 repository · arXiv:2003.02692
-
Big Transfer (BiT): General Visual Representation Learning 24 Dec 2019 · 9 repositories · arXiv:1912.11370Syntology community repositories only · 9 ran (of which 0 constructed an object rather than computing a result; 7 with no instrument failure: 0 honoured, 1 violated, 6 with no contract checked; 2 where Syntology's instrument failed) · 1 unverified (of 10 harvested samples)
-
PointRend: Image Segmentation as Rendering 17 Dec 2019 · 14 repositories · arXiv:1912.08193Syntology community repositories only · 16 ran (of which 0 constructed an object rather than computing a result; 11 with no instrument failure: 0 honoured, 0 violated, 11 with no contract checked; 5 where Syntology's instrument failed) · 2 unverified (of 18 harvested samples)
-
Local Context Normalization: Revisiting Local Normalization 12 Dec 2019 · 1 repository · arXiv:1912.05845
-
Bridging the Gap Between Anchor-based and Anchor-free Detection via Adaptive Training Sample Selection 5 Dec 2019 · 13 repositories · arXiv:1912.02424Syntology community repositories only · 4 ran (of which 0 constructed an object rather than computing a result; 4 with no instrument failure: 0 honoured, 0 violated, 4 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified (of 4 harvested samples)
-
An Exponential Learning Rate Schedule for Deep Learning 16 Oct 2019 · 0 repositories · arXiv:1910.07454
-
The Non-IID Data Quagmire of Decentralized Machine Learning 1 Oct 2019 · 1 repository · arXiv:1910.00189
-
An Empirical Study of Batch Normalization and Group Normalization in Conditional Computation 31 Jul 2019 · 0 repositories · arXiv:1908.00061
-
Switchable Normalization for Learning-to-Normalize Deep Representation 22 Jul 2019 · 0 repositories · arXiv:1907.10473
-
IMP: Instance Mask Projection for High Accuracy Semantic Segmentation of Things 15 Jun 2019 · 0 repositories · arXiv:1906.06597
-
Four Things Everyone Should Know to Improve Batch Normalization 9 Jun 2019 · 2 repositories · arXiv:1906.03548Syntology official (archive's flag): 1 ran · 1 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 1 honoured, 0 violated, 0 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified (of 1 harvested sample) · 1 pointer-only (licence)
-
Online Normalization for Training Neural Networks 15 May 2019 · 1 repository · arXiv:1905.05894Syntology official (archive's flag): 9 ran · 9 ran (of which 0 constructed an object rather than computing a result; 9 with no instrument failure: 0 honoured, 0 violated, 9 with no contract checked; 0 where Syntology's instrument failed) · 7 unverified (of 16 harvested samples)
-
Instance-Level Meta Normalization 6 Apr 2019 · 1 repository · arXiv:1904.03516
-
RetinaMask: Learning to predict masks improves state-of-the-art single-shot detection for free 10 Jan 2019 · 53 repositories · arXiv:1901.03353
-
Panoptic Feature Pyramid Networks 8 Jan 2019 · 12 repositories · arXiv:1901.02446Syntology community repositories only · 11 ran (of which 0 constructed an object rather than computing a result; 4 with no instrument failure: 0 honoured, 0 violated, 4 with no contract checked; 7 where Syntology's instrument failed) · 1 unverified (of 12 harvested samples) · 2 pointer-only (licence)
-
Kalman Normalization: Normalizing Internal Representations Across Network Layers 1 Dec 2018 · 0 repositories
-
Normalization in Training U-Net for 2D Biomedical Semantic Segmentation 11 Sep 2018 · 1 repository · arXiv:1809.03783
-
Differentiable Learning-to-Normalize via Switchable Normalization 28 Jun 2018 · 3 repositories · arXiv:1806.10779Syntology 10 ran (of which 0 constructed an object rather than computing a result; 10 with no instrument failure: 0 honoured, 0 violated, 10 with no contract checked; 0 where Syntology's instrument failed) · 3 unverified (of 13 harvested samples)
-
Group Normalization 22 Mar 2018 · 22 repositories · arXiv:1803.08494Syntology official (archive's flag): 3 ran · 7 ran (of which 2 constructed an object rather than computing a result; 5 with no instrument failure: 0 honoured, 0 violated, 5 with no contract checked; 2 where Syntology's instrument failed) · 8 unverified (of 15 harvested samples) · 5 pointer-only (licence)
-
Cascade R-CNN: Delving into High Quality Object Detection 3 Dec 2017 · 8 repositories · arXiv:1712.00726Syntology official: no sample here; runs from other or unrecorded repositories · 2 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 2 where Syntology's instrument failed) · 0 unverified (of 2 harvested samples) · 2 pointer-only (licence)
Tasks archive 2025-07-28
20 shown of 72 tasks the archive attaches to papers tagged with this method, by distinct papers. A task without a page in the catalog is plain text.
Usage over time archive 2025-07-28
Components: the archive holds no method-to-method composition, so PwC's Components table cannot be rebuilt; the Papers list carries no Results column for the same reason (the archive does not join its leaderboard rows to method tags).
Categories archive 2025-07-28
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