Hybrid Focal and Full-Range Attention Based Graph Transformers

• URL: http://arxiv.org/abs/2311.04653v2
• Date: Tue, 10 Sep 2024 03:38:37 GMT
• Title: Hybrid Focal and Full-Range Attention Based Graph Transformers
• Authors: Minhong Zhu, Zhenhao Zhao, Weiran Cai,
• Abstract summary: We present a purely attention-based architecture, namely Focal and Full-Range Graph Transformer (FFGT) FFGT combines the conventional full-range attention with K-hop focal attention on ego-nets to aggregate both global and local information. Our approach enhances the performance of existing Graph Transformers on various open datasets.
• Score: 0.0
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: The paradigm of Transformers using the self-attention mechanism has manifested its advantage in learning graph-structured data. Yet, Graph Transformers are capable of modeling full range dependencies but are often deficient in extracting information from locality. A common practice is to utilize Message Passing Neural Networks (MPNNs) as an auxiliary to capture local information, which however are still inadequate for comprehending substructures. In this paper, we present a purely attention-based architecture, namely Focal and Full-Range Graph Transformer (FFGT), which can mitigate the loss of local information in learning global correlations. The core component of FFGT is a new mechanism of compound attention, which combines the conventional full-range attention with K-hop focal attention on ego-nets to aggregate both global and local information. Beyond the scope of canonical Transformers, the FFGT has the merit of being more substructure-aware. Our approach enhances the performance of existing Graph Transformers on various open datasets, while achieves compatible SOTA performance on several Long-Range Graph Benchmark (LRGB) datasets even with a vanilla transformer. We further examine influential factors on the optimal focal length of attention via introducing a novel synthetic dataset based on SBM-PATTERN.

Related papers

• CARE Transformer: Mobile-Friendly Linear Visual Transformer via Decoupled Dual Interaction [77.8576094863446]
  We propose a new detextbfCoupled dutextbfAl-interactive lineatextbfR atttextbfEntion (CARE) mechanism. We first propose an asymmetrical feature decoupling strategy that asymmetrically decouples the learning process for local inductive bias and long-range dependencies. By adopting a decoupled learning way and fully exploiting complementarity across features, our method can achieve both high efficiency and accuracy.
  arXiv  Detail & Related papers  (2024-11-25T07:56:13Z)
• SGFormer: Single-Layer Graph Transformers with Approximation-Free Linear Complexity [74.51827323742506]
  We evaluate the necessity of adopting multi-layer attentions in Transformers on graphs. We show that one-layer propagation can be reduced to one-layer propagation, with the same capability for representation learning. It suggests a new technical path for building powerful and efficient Transformers on graphs.
  arXiv  Detail & Related papers  (2024-09-13T17:37:34Z)
• Less is More: on the Over-Globalizing Problem in Graph Transformers [34.52455014631614]
  The global attention mechanism considers a wider receptive field in a fully connected graph, leading many to believe that useful information can be extracted from all the nodes. We show that the current attention mechanism overly focuses on those distant nodes, while the near nodes, which actually contain most of the useful information, are relatively weakened. We propose a novel Bi-Level Global Graph Transformer with Collaborative Training (CoBFormer) to prevent the over-globalizing problem.
  arXiv  Detail & Related papers  (2024-05-02T09:12:22Z)
• DAT++: Spatially Dynamic Vision Transformer with Deformable Attention [87.41016963608067]
  We present Deformable Attention Transformer ( DAT++), a vision backbone efficient and effective for visual recognition. DAT++ achieves state-of-the-art results on various visual recognition benchmarks, with 85.9% ImageNet accuracy, 54.5 and 47.0 MS-COCO instance segmentation mAP, and 51.5 ADE20K semantic segmentation mIoU.
  arXiv  Detail & Related papers  (2023-09-04T08:26:47Z)
• SGFormer: Simplifying and Empowering Transformers for Large-Graph Representations [75.71298846760303]
  We show that a one-layer attention can bring up surprisingly competitive performance across node property prediction benchmarks. We frame the proposed scheme as Simplified Graph Transformers (SGFormer), which is empowered by a simple attention model. We believe the proposed methodology alone enlightens a new technical path of independent interest for building Transformers on large graphs.
  arXiv  Detail & Related papers  (2023-06-19T08:03:25Z)
• Are More Layers Beneficial to Graph Transformers? [97.05661983225603]
  Current graph transformers suffer from the bottleneck of improving performance by increasing depth. Deep graph transformers are limited by the vanishing capacity of global attention. We propose a novel graph transformer model named DeepGraph that explicitly employs substructure tokens in the encoded representation.
  arXiv  Detail & Related papers  (2023-03-01T15:22:40Z)
• Edge-augmented Graph Transformers: Global Self-attention is Enough for Graphs [24.796242917673755]
  We propose a simple yet powerful extension to the transformer - residual edge channels. The resultant framework, which we call Edge-augmented Graph Transformer (EGT), can directly accept, process and output structural information as well as node information. Our framework, which relies on global node feature aggregation, achieves better performance compared to Graph Convolutional Networks (GCN)
  arXiv  Detail & Related papers  (2021-08-07T02:18:11Z)
• Feature Fusion Vision Transformer for Fine-Grained Visual Categorization [22.91753200323264]
  We propose a novel pure transformer-based framework Feature Fusion Vision Transformer (FFVT) We aggregate the important tokens from each transformer layer to compensate the local, low-level and middle-level information. We design a novel token selection mod-ule called mutual attention weight selection (MAWS) to guide the network effectively and efficiently towards selecting discriminative tokens.
  arXiv  Detail & Related papers  (2021-07-06T01:48:43Z)
• Rethinking Graph Transformers with Spectral Attention [13.068288784805901]
  We present the $textitSpectral Attention Network$ (SAN), which uses a learned positional encoding (LPE) to learn the position of each node in a given graph. By leveraging the full spectrum of the Laplacian, our model is theoretically powerful in distinguishing graphs, and can better detect similar sub-structures from their resonance. Our model performs on par or better than state-of-the-art GNNs, and outperforms any attention-based model by a wide margin.
  arXiv  Detail & Related papers  (2021-06-07T18:11:11Z)
• Multi-Level Graph Convolutional Network with Automatic Graph Learning for Hyperspectral Image Classification [63.56018768401328]
  We propose a Multi-level Graph Convolutional Network (GCN) with Automatic Graph Learning method (MGCN-AGL) for HSI classification. By employing attention mechanism to characterize the importance among spatially neighboring regions, the most relevant information can be adaptively incorporated to make decisions. Our MGCN-AGL encodes the long range dependencies among image regions based on the expressive representations that have been produced at local level.
  arXiv  Detail & Related papers  (2020-09-19T09:26:20Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.