Community Implementations: [![CatalyzeX](/images/catalyzex_icon.svg) 1 code implementation](https://www.catalyzex.com/paper/arxiv:2008.04428/code)
Track: full conference paper
TL;DR: We propose an image pyramid based approach to regressing landmark locations in large images efficiently with a cnn via foveation (non-uniform sampling).
Keywords: Deep learning, Landmark detection, Attention mechanism, Convolutional Neural Network, 2D X-ray cephalometric analysis, Image pyramid
Abstract: CNNs, initially inspired by human vision, differ in a key way: they sample uniformly, rather than with highest density in a focal point. For very large images, this makes training untenable, as the memory and computation required for activation maps scales quadratically with the side length of an image. We propose an image pyramid based approach that extracts narrow glimpses of the of the input image and iteratively refines them to accomplish regression tasks. To assist with high-accuracy regression, we introduce a novel intermediate representation we call ‘spatialized features’. Our approach scales logarithmically with the side length, so it works with very large images. We apply our method to Cephalometric X-ray Landmark Detection and get state-of-the-art results.
Paper Type: methodological development
Source Latex: zip
Presentation Upload: zip
Presentation Upload Agreement: I agree that my presentation material (videos and slides) will be made public.