Meta Platforms, Inc.
GENERATING REFINED OBJECT PROPOSALS USING DEEP-LEARNING MODELS

Abstract:

In one embodiment, a feature map of an image having h.times.w pixels and a patch having one or more pixels of the image are received. The patch has been processed by a first set of layers of a convolutional neural network and contains an object centered within the patch. The patch is then processed using the feature map and one or more pixel classifiers of a classification layer of a deep-learning model, where the classification layer includes h.times.w pixel classifiers, with each pixel classifier corresponding to a respective pixel of the patch. Each of the pixel classifiers used to process the patch outputs a respective value indicating whether the corresponding pixel belongs to the object centered in the patch.