Methods › Computer Vision › Image Models › CvT
Convolutional Vision Transformer
CvT
Introduced by Haiping Wu et al. in CvT: Introducing Convolutions to Vision Transformers
archive 2025-07-28 Description, source and code snippet are the archive's method entry.
The Convolutional vision Transformer (CvT) is an architecture which incorporates convolutions into the Transformer. The CvT design introduces convolutions to two core sections of the ViT architecture.
First, the Transformers are partitioned into multiple stages that form a hierarchical structure of Transformers. The beginning of each stage consists of a convolutional token embedding that performs an overlapping convolution operation with stride on a 2D-reshaped token map (i.e., reshaping flattened token sequences back to the spatial grid), followed by layer normalization. This allows the model to not only capture local information, but also progressively decrease the sequence length while simultaneously increasing the dimension of token features across stages, achieving spatial downsampling while concurrently increasing the number of feature maps, as is performed in CNNs.
Second, the linear projection prior to every self-attention block in the Transformer module is replaced with a proposed convolutional projection, which employs a s × s depth-wise separable convolution operation on an 2D-reshaped token map. This allows the model to further capture local spatial context and reduce semantic ambiguity in the attention mechanism. It also permits management of computational complexity, as the stride of convolution can be used to subsample the key and value matrices to improve efficiency by 4× or more, with minimal degradation of performance.
Papers archive 2025-07-28
12 shown of 12, 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.
-
Residual Transformer Fusion Network for Salt and Pepper Image Denoising 13 Feb 2025 · 0 repositories · arXiv:2502.09000
-
Heracles: A Hybrid SSM-Transformer Model for High-Resolution Image and Time-Series Analysis 26 Mar 2024 · 2 repositories · arXiv:2403.18063Syntology ran 4 of 4 samples · 0 unverified · 4 pointer-only (licence)
-
A Comparison of Deep Learning Models for Proton Background Rejection with the AMS Electromagnetic Calorimeter 26 Feb 2024 · 0 repositories · arXiv:2402.16285
-
On Convolutional Vision Transformers for Yield Prediction 8 Feb 2024 · 0 repositories · arXiv:2402.05557
-
Making Vision Transformers Truly Shift-Equivariant 25 May 2023 · 0 repositories · arXiv:2305.16316
-
A Novel Framework for Decentralized Dynamic Resource Allocation Using Voronoi Tessellations 30 Nov 2022 · 0 repositories · arXiv:2212.00140
-
Online Continual Learning with Contrastive Vision Transformer 24 Jul 2022 · 0 repositories · arXiv:2207.13516
-
SimA: Simple Softmax-free Attention for Vision Transformers 17 Jun 2022 · 1 repository · arXiv:2206.08898Syntology ran 0 of 3 samples · 3 unverified
-
Convolutional Xformers for Vision 25 Jan 2022 · 1 repository · arXiv:2201.10271
-
Energy-optimal Design and Control of Electric Powertrains under Motor Thermal Constraints 15 Nov 2021 · 0 repositories · arXiv:2111.07711
-
Convolutional Gated MLP: Combining Convolutions & gMLP 6 Nov 2021 · 0 repositories · arXiv:2111.03940
-
CvT: Introducing Convolutions to Vision Transformers 29 Mar 2021 · 16 repositories · arXiv:2103.15808Syntology ran 29 of 47 samples · 18 unverified · 4 pointer-only (licence)
Tasks archive 2025-07-28
18 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