Methods › General › Attention Mechanisms › Deformable ConvNets
Deformable Convolutional Networks
Deformable ConvNets
Introduced by Jifeng Dai et al. in Deformable Convolutional Networks
archive 2025-07-28 Description, source and code snippet are the archive's method entry.
Deformable ConvNets do not learn an affine transformation. They divide convolution into two steps, firstly sampling features on a regular grid ℛ from the input feature map, then aggregating sampled features by weighted summation using a convolution kernel. The process can be written as: Y(p₀) = ∑_(pᵢ ∈ℛ) w(pᵢ) X(p₀ + pᵢ) ℛ = {(-1,-1), (-1, 0), …, (1, 1)} The deformable convolution augments the sampling process by introducing a group of learnable offsets Δpᵢ which can be generated by a lightweight CNN. Using the offsets Δpᵢ, the deformable convolution can be formulated as: Y(p₀) = ∑_(pᵢ ∈ℛ) w(pᵢ) X(p₀ + pᵢ + Δpᵢ). Through the above method, adaptive sampling is achieved. However, Δpᵢ is a floating point value unsuited to grid sampling. To address this problem, bilinear interpolation is used. Deformable RoI pooling is also used, which greatly improves object detection.
Deformable ConvNets adaptively select the important regions and enlarge the valid receptive field of convolutional neural networks; this is important in object detection and semantic segmentation tasks.
Papers archive 2025-07-28
2 shown of 2, newest first. Repository counts are the archive's code-links table. A Syntology line states what Syntology ran from that paper's harvested code; it is per sample and not a correctness claim.
-
Deformable ConvNets v2: More Deformable, Better Results 27 Nov 2018 · 26 repositories · arXiv:1811.11168Syntology ran 2 of 13 samples · 11 unverified · 2 pointer-only (licence)
-
Deformable Convolutional Networks 17 Mar 2017 · 38 repositories · arXiv:1703.06211Syntology ran 3 of 9 samples · 6 unverified · 3 pointer-only (licence)
Tasks archive 2025-07-28
5 tasks the archive attaches to papers tagged with this method, by distinct papers. A task without a page in the catalog is plain text.
| Task | Papers |
|---|---|
| Object Detection | 2 |
| Semantic Segmentation | 2 |
| Instance Segmentation | 1 |
| Object | 1 |
| Vessel Detection | 1 |
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