Explicitly modeled attention maps for image classification

A. Tan, Tam Nguyen, M. Dax, M. Nießner, Thomas Brox
AAAI Conference on Artificial Intelligence, 2021
Abstract: Self-attention networks have shown remarkable progress in computer vision tasks such as image classification. The main benefit of the self-attention mechanism is the ability to capture long-range feature interactions in attention-maps. However, the computation of attention-maps requires a learnable key, query, and positional encoding, whose usage is often not intuitive and computationally expensive. To mitigate this problem, we propose a novel self-attention module with explicitly modeled attention-maps using only a single learnable parameter for low computational overhead. The design of explicitly modeled attention-maps using a geometric prior is based on the observation that the spatial context for a given pixel within an image is mostly dominated by its neighbors, while more distant pixels have a minor contribution. Concretely, the attention-maps are parametrized via simple functions (e.g., Gaussian kernel) with a learnable radius, which is modeled independently of the input content. Our evaluation shows that our method achieves an accuracy improvement of up to 2.2% over the ResNet-baselines in ImageNet ILSVRC and outperforms other self-attention methods such as AA-ResNet152 in accuracy by 0.9% with 6.4% fewer parameters and 6.7% fewer GFLOPs. This result empirically indicates the value of incorporating a geometric prior into the self-attention mechanism when applied in image classification.

Other associated files : AAAImain.pdf [1.7MB]  

Images and movies


BibTex reference

  author       = "A. Tan and D. T. Nguyen and M. Dax and M. Nießner and T. Brox",
  title        = "Explicitly modeled attention maps for image classification",
  booktitle    = "AAAI Conference on Artificial Intelligence",
  month        = " ",
  year         = "2021",
  url          = "http://lmb.informatik.uni-freiburg.de/Publications/2021/NB21"

Other publications in the database