Autofocus Layer for Semantic Segmentation
This addresses multi-scale feature extraction for medical image segmentation, though it appears incremental as it builds on existing attention and dilation techniques.
The authors tackled the problem of multi-scale processing in semantic segmentation by proposing an autofocus convolutional layer that adaptively adjusts receptive field size based on context, achieving very promising performance on pelvic CT and brain tumor MRI segmentation tasks.
We propose the autofocus convolutional layer for semantic segmentation with the objective of enhancing the capabilities of neural networks for multi-scale processing. Autofocus layers adaptively change the size of the effective receptive field based on the processed context to generate more powerful features. This is achieved by parallelising multiple convolutional layers with different dilation rates, combined by an attention mechanism that learns to focus on the optimal scales driven by context. By sharing the weights of the parallel convolutions we make the network scale-invariant, with only a modest increase in the number of parameters. The proposed autofocus layer can be easily integrated into existing networks to improve a model's representational power. We evaluate our models on the challenging tasks of multi-organ segmentation in pelvic CT and brain tumor segmentation in MRI and achieve very promising performance.