Methods › General › Attention Modules › Deformable Attention Module
Deformable Attention Module
Introduced by Xizhou Zhu et al. in Deformable DETR: Deformable Transformers for End-to-End Object Detection
archive 2025-07-28 Description, source and code snippet are the archive's method entry.
Deformable Attention Module is an attention module used in the Deformable DETR architecture, which seeks to overcome one issue base Transformer attention in that it looks over all possible spatial locations. Inspired by deformable convolution, the deformable attention module only attends to a small set of key sampling points around a reference point, regardless of the spatial size of the feature maps. By assigning only a small fixed number of keys for each query, the issues of convergence and feature spatial resolution can be mitigated.
Given an input feature map x ∈ℝ^(C ×H ×W), let q index a query element with content feature z_q and a 2-d reference point p_q, the deformable attention feature is calculated by:
DeformAttn(z_q, p_q, x)=∑ₘ₌₁ᴹ Wₘ[∑ₖ₌₁ᴷ A_(m q k) ·Wₘ′ x(p_q+Δp_(m q k))]
where m indexes the attention head, k indexes the sampled keys, and K is the total sampled key number (K ≪H W) . Δp_(m q k) and A_(m q k) denote the sampling offset and attention weight of the kᵗʰ sampling point in the mᵗʰ attention head, respectively. The scalar attention weight A_(m q k) lies in the range [0,1], normalized by ∑ₖ₌₁ᴷ A_(m q k)=1 . Δp_(m q k) ∈ℝ² are of 2-d real numbers with unconstrained range. As p_q+Δp_(m q k) is fractional, bilinear interpolation is applied as in Dai et al. (2017) in computing x(p_q+Δp_(m q k)). Both Δp_(m q k) and A_(m q k) are obtained via linear projection over the query feature z_q . In implementation, the query feature z_q is fed to a linear projection operator of 3 M K channels, where the first 2 M K channels encode the sampling offsets Δp_(m q k), and the remaining M K channels are fed to a softmax operator to obtain the attention weights A_(m q k).
In Syntology name not resolved (unknown_name). The lookup matched no method, paper or venue by this name; that asserts nothing about whether code exists.
Near matches the lookup lists, each a different name: part-aware deformable attention; Deformable Attention Transformer; Deformable Attention Flow (DAFlow); deformable cross attention mechanism; Multi-Scale Deformable Attention (MSDA); Reference-based Deformable Self-Attention; EL-Attention; TF-ATTENTION.
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.
Papers archive 2025-07-28
archive papers tagged: 42 · with a code link: 25 · where Syntology ran a sample: 9 (8 with a run with no instrument failure, 1 where every run was a failure of Syntology's instrument) Syntology
Show: all tagged papersonly where code ran (9 of 42 tagged: 8 with a run with no instrument failure, 1 where every run was a failure of Syntology's instrument)
Page 1 of 1: papers 1 to 42 of 42, 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.
-
WalnutData: A UAV Remote Sensing Dataset of Green Walnuts and Model Evaluation 27 Feb 2025 · 1 repository · arXiv:2502.20092
-
Advancing SEM Based Nano-Scale Defect Analysis in Semiconductor Manufacturing for Advanced IC Nodes 6 Sep 2024 · 0 repositories · arXiv:2409.04310
-
U-DECN: End-to-End Underwater Object Detection ConvNet with Improved DeNoising Training 11 Aug 2024 · 1 repository · arXiv:2408.05780
-
Fisher-aware Quantization for DETR Detectors with Critical-category Objectives 3 Jul 2024 · 0 repositories · arXiv:2407.03442
-
Knowledge-driven Subspace Fusion and Gradient Coordination for Multi-modal Learning 20 Jun 2024 · 1 repository · arXiv:2406.13979
-
Understanding differences in applying DETR to natural and medical images 27 May 2024 · 0 repositories · arXiv:2405.17677
-
Infrared Adversarial Car Stickers 16 May 2024 · 0 repositories · arXiv:2405.09924
-
LDTR: Transformer-based Lane Detection with Anchor-chain Representation 21 Mar 2024 · 0 repositories · arXiv:2403.14354
-
Generative Region-Language Pretraining for Open-Ended Object Detection 15 Mar 2024 · 1 repository · arXiv:2403.10191Syntology official (archive's flag): 8 ran · 8 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; 6 where Syntology's instrument failed) · 3 unverified (of 11 harvested samples) · 4 pointer-only (licence)
-
Hybrid Proposal Refiner: Revisiting DETR Series from the Faster R-CNN Perspective 1 Jan 2024 · 1 repository
-
KD-DETR: Knowledge Distillation for Detection Transformer with Consistent Distillation Points Sampling 1 Jan 2024 · 0 repositories
-
Mono3DVG: 3D Visual Grounding in Monocular Images 13 Dec 2023 · 1 repository · arXiv:2312.08022Syntology 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) · 1 unverified (of 7 harvested samples) · 7 pointer-only (licence)
-
Towards Few-Annotation Learning for Object Detection: Are Transformer-based Models More Efficient ? 30 Oct 2023 · 1 repository · arXiv:2310.19936
-
DAC-DETR: Divide the Attention Layers and Conquer 21 Sep 2023 · 1 repository
-
A Spatial-Temporal Deformable Attention based Framework for Breast Lesion Detection in Videos 9 Sep 2023 · 1 repository · arXiv:2309.04702
-
Selecting Learnable Training Samples is All DETRs Need in Crowded Pedestrian Detection 18 May 2023 · 0 repositories · arXiv:2305.10801
-
Robust Traffic Light Detection Using Salience-Sensitive Loss: Computational Framework and Evaluations 8 May 2023 · 0 repositories · arXiv:2305.04516
-
Continual Detection Transformer for Incremental Object Detection 6 Apr 2023 · 0 repositories · arXiv:2304.03110
-
VDDT: Improving Vessel Detection with Deformable Transfomer 15 Mar 2023 · 0 repositories
-
ARS-DETR: Aspect Ratio-Sensitive Detection Transformer for Aerial Oriented Object Detection 9 Mar 2023 · 1 repository · arXiv:2303.04989
-
Towards Robust Video Instance Segmentation with Temporal-Aware Transformer 20 Jan 2023 · 0 repositories · arXiv:2301.09416
-
Salient Sign Detection In Safe Autonomous Driving: AI Which Reasons Over Full Visual Context 14 Jan 2023 · 0 repositories · arXiv:2301.05804
-
Cross-Modal Learning with 3D Deformable Attention for Action Recognition 12 Dec 2022 · 0 repositories · arXiv:2212.05638
-
Open World DETR: Transformer based Open World Object Detection 6 Dec 2022 · 0 repositories · arXiv:2212.02969
-
FQDet: Fast-converging Query-based Detector 5 Oct 2022 · 2 repositories · arXiv:2210.02318Syntology official (archive's flag): 2 ran · 3 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; 1 where Syntology's instrument failed) · 3 unverified (of 6 harvested samples) · 3 pointer-only (licence)
-
ComplETR: Reducing the cost of annotations for object detection in dense scenes with vision transformers 13 Sep 2022 · 0 repositories · arXiv:2209.05654
-
DPText-DETR: Towards Better Scene Text Detection with Dynamic Points in Transformer 10 Jul 2022 · 1 repository · arXiv:2207.04491
-
YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors 6 Jul 2022 · 21 repositories · arXiv:2207.02696Syntology community repositories only · 10 ran (of which 0 constructed an object rather than computing a result; 8 with no instrument failure: 2 honoured, 0 violated, 6 with no contract checked; 2 where Syntology's instrument failed) · 1 unverified (of 11 harvested samples) · 5 pointer-only (licence)
-
An Empirical Study Of Self-supervised Learning Approaches For Object Detection With Transformers 11 May 2022 · 2 repositories · arXiv:2205.05543
-
MTTrans: Cross-Domain Object Detection with Mean-Teacher Transformer 3 May 2022 · 1 repository · arXiv:2205.01643
-
DFAM-DETR: Deformable feature based attention mechanism DETR on slender object detection 22 Apr 2022 · 0 repositories · arXiv:2204.10667
-
Deformable VisTR: Spatio temporal deformable attention for video instance segmentation 12 Mar 2022 · 1 repository · arXiv:2203.06318
-
Task Specific Attention is one more thing you need for object detection 18 Feb 2022 · 1 repository · arXiv:2202.09048
-
TransVOD: End-to-End Video Object Detection with Spatial-Temporal Transformers 13 Jan 2022 · 3 repositories · arXiv:2201.05047Syntology official (archive's flag): 3 ran · 6 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 1 violated, 2 with no contract checked; 3 where Syntology's instrument failed) · 0 unverified (of 6 harvested samples) · 3 pointer-only (licence)
-
Recurrent Glimpse-based Decoder for Detection with Transformer 9 Dec 2021 · 1 repository · arXiv:2112.04632
-
Sparse DETR: Efficient End-to-End Object Detection with Learnable Sparsity 29 Nov 2021 · 1 repository · arXiv:2111.14330Syntology official (archive's flag): 3 ran · 3 ran (of which 3 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 0 violated, 3 with no contract checked; 0 where Syntology's instrument failed) · 4 unverified; every one of the 3 samples that ran constructed an object rather than computing a result (of 7 harvested samples)
-
Panoptic SegFormer: Delving Deeper into Panoptic Segmentation with Transformers 8 Sep 2021 · 3 repositories · arXiv:2109.03814
-
End-to-end Temporal Action Detection with Transformer 18 Jun 2021 · 1 repository · arXiv:2106.10271Syntology official (archive's flag): 2 ran · 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) · 1 unverified (of 3 harvested samples)
-
DETReg: Unsupervised Pretraining with Region Priors for Object Detection 8 Jun 2021 · 1 repository · arXiv:2106.04550Syntology official (archive's flag): 5 ran · 5 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 1 violated, 1 with no contract checked; 3 where Syntology's instrument failed) · 2 unverified (of 7 harvested samples) · 1 pointer-only (licence)
-
End-to-End Video Object Detection with Spatial-Temporal Transformers 23 May 2021 · 1 repository · arXiv:2105.10920
-
Efficient DETR: Improving End-to-End Object Detector with Dense Prior 3 Apr 2021 · 0 repositories · arXiv:2104.01318
-
Deformable DETR: Deformable Transformers for End-to-End Object Detection 8 Oct 2020 · 20 repositories · arXiv:2010.04159Syntology community repositories only · 37 ran (of which 9 constructed an object rather than computing a result; 24 with no instrument failure: 1 honoured, 3 violated, 20 with no contract checked; 13 where Syntology's instrument failed) · 18 unverified (of 55 harvested samples) · 21 pointer-only (licence)
Tasks archive 2025-07-28
20 shown of 66 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