Title :
Labeling Spain With Stanford
Author :
Yingbo Zhou ; Nwogu, Ifeoma ; Govindaraju, Vengatesan
Author_Institution :
Dept. of Comput. Sci. & Eng., State Univ. of New York at Buffalo, Buffalo, NY, USA
Abstract :
We present an end-to-end framework for outdoor scene region decomposition, learned on a small set of randomly selected images that generalizes well to multiple data sets containing images from around the world. We discuss the different aspects of the framework especially a generalized variational inference method with better approximations to the true marginals of a graphical model. Experimentally, we explain why the framework is robust and performs competitively on many diverse scene data sets, including several unseen scene types. We have obtained high pixel-level accuracies ( ≈ 80%) in three of the four data sets, which include a benchmark data set known as the Stanford background data set. Our model obtained over 70% accuracy on the fourth data set, which contained a number of indoor and close-up images that are significantly different from our training examples.
Keywords :
computer vision; image segmentation; inference mechanisms; Spain; Stanford background data set; generalized variational inference method; graphical model; multiple data set; outdoor scene region decomposition; Accuracy; Approximation methods; Benchmark testing; Clustering algorithms; Image color analysis; Image segmentation; Training; Scene understanding; generalization; generalized mean field; low- and mid-level image cues; semantic labeling;
Journal_Title :
Image Processing, IEEE Transactions on
DOI :
10.1109/TIP.2013.2285603