Cross-Layer Retrospective Retrieving via Layer Attention
This work addresses the need for more powerful vision networks for researchers and practitioners, offering a novel method that is incremental but shows strong gains in specific applications.
The paper tackles the problem of enhancing deep neural network representation by strengthening layer interactions, introducing a cross-layer attention mechanism called MRLA that improves performance across vision tasks, achieving a 1.6% Top-1 accuracy gain on ResNet-50 and 3-4% AP boosts in dense prediction tasks.
More and more evidence has shown that strengthening layer interactions can enhance the representation power of a deep neural network, while self-attention excels at learning interdependencies by retrieving query-activated information. Motivated by this, we devise a cross-layer attention mechanism, called multi-head recurrent layer attention (MRLA), that sends a query representation of the current layer to all previous layers to retrieve query-related information from different levels of receptive fields. A light-weighted version of MRLA is also proposed to reduce the quadratic computation cost. The proposed layer attention mechanism can enrich the representation power of many state-of-the-art vision networks, including CNNs and vision transformers. Its effectiveness has been extensively evaluated in image classification, object detection and instance segmentation tasks, where improvements can be consistently observed. For example, our MRLA can improve 1.6% Top-1 accuracy on ResNet-50, while only introducing 0.16M parameters and 0.07B FLOPs. Surprisingly, it can boost the performances by a large margin of 3-4% box AP and mask AP in dense prediction tasks. Our code is available at https://github.com/joyfang1106/MRLA.