Multi-scale gridded Gabor attention for cirrus segmentation
This addresses the challenge of efficiently segmenting contaminants in large-scale images, such as astronomical data, with incremental improvements in attention mechanisms.
The paper tackles the problem of segmenting global contaminants in large images by proposing a gridded attention mechanism that improves efficiency and texture sensitivity, achieving results on a new astronomical dataset for dust cloud segmentation.
In this paper, we address the challenge of segmenting global contaminants in large images. The precise delineation of such structures requires ample global context alongside understanding of textural patterns. CNNs specialise in the latter, though their ability to generate global features is limited. Attention measures long range dependencies in images, capturing global context, though at a large computational cost. We propose a gridded attention mechanism to address this limitation, greatly increasing efficiency by processing multi-scale features into smaller tiles. We also enhance the attention mechanism for increased sensitivity to texture orientation, by measuring correlations across features dependent on different orientations, in addition to channel and positional attention. We present results on a new dataset of astronomical images, where the task is segmenting large contaminating dust clouds.