Feedforward semantic segmentation with zoom-out features
This work addresses semantic segmentation for computer vision applications, presenting a novel feed-forward approach that avoids complex inference mechanisms.
The paper tackles semantic segmentation by classifying superpixels using a feed-forward network that extracts features from nested regions of increasing extent, achieving 64.4% average accuracy on the PASCAL VOC 2012 test set.
We introduce a purely feed-forward architecture for semantic segmentation. We map small image elements (superpixels) to rich feature representations extracted from a sequence of nested regions of increasing extent. These regions are obtained by "zooming out" from the superpixel all the way to scene-level resolution. This approach exploits statistical structure in the image and in the label space without setting up explicit structured prediction mechanisms, and thus avoids complex and expensive inference. Instead superpixels are classified by a feedforward multilayer network. Our architecture achieves new state of the art performance in semantic segmentation, obtaining 64.4% average accuracy on the PASCAL VOC 2012 test set.