Segaware

show more

Implementations

Segmentation-Aware Convolutional Networks Using Local Attention Masks

Segmentation-aware convolution filters are invariant to backgrounds. We achieve this in three steps: (i) compute segmentation cues for each pixel (i.e., “embeddings”), (ii) create a foreground mask for each patch, and (iii) combine the masks with convolution, so that the filters only process the local foreground in each image patch.

Github: https://github.com/aharley/segaware

segmentationsegaware
How helpful was this page?