Aggregating Global Features into Local Vision Transformer
This work addresses a specific bottleneck in vision Transformers for computer vision tasks, offering an incremental improvement in model efficiency and accuracy.
The paper tackles the problem of unclear global information aggregation in local window-based vision Transformers by proposing a multi-resolution overlapped attention (MOA) module, which leads to significant performance gains and outperforms previous vision Transformers with fewer parameters on datasets like CIFAR-10, CIFAR-100, and ImageNet-1K.
Local Transformer-based classification models have recently achieved promising results with relatively low computational costs. However, the effect of aggregating spatial global information of local Transformer-based architecture is not clear. This work investigates the outcome of applying a global attention-based module named multi-resolution overlapped attention (MOA) in the local window-based transformer after each stage. The proposed MOA employs slightly larger and overlapped patches in the key to enable neighborhood pixel information transmission, which leads to significant performance gain. In addition, we thoroughly investigate the effect of the dimension of essential architecture components through extensive experiments and discover an optimum architecture design. Extensive experimental results CIFAR-10, CIFAR-100, and ImageNet-1K datasets demonstrate that the proposed approach outperforms previous vision Transformers with a comparatively fewer number of parameters.