2015 · 190 citations · 24 references
ComputingComputer ScienceComputer VisionMachine LearningRendering3D ReconstructionDeep LearningImage SynthesisPose EstimationScene ReconstructionScene Understanding3D Pose Estimation6D Object Pose Estimation6D Pose EstimationConvolutional Neural NetworksMaximum Likelihood InferenceObject Pose EstimationObjectsOcclusion Compensation
Analysis‑by‑synthesis has proven effective for 6D pose estimation in RGB‑D images, but occlusion and sensor noise make meaningful comparison difficult. The authors propose a method that learns to compare observations with rendered images, thereby addressing occlusion and noise challenges. They model the posterior pose density with a CNN that compares observed and rendered images, trained via maximum likelihood. Empirically, the CNN generalizes across diverse objects and backgrounds, achieving significant improvements over state‑of‑the‑art on two datasets with eleven objects, clutter, and heavy occlusion.
Analysis-by-synthesis has been a successful approach for many tasks in computer vision, such as 6D pose estimation of an object in an RGB-D image which is the topic of this work. The idea is to compare the observation with the output of a forward process, such as a rendered image of the object of interest in a particular pose. Due to occlusion or complicated sensor noise, it can be difficult to perform this comparison in a meaningful way. We propose an approach that "learns to compare", while taking these difficulties into account. This is done by describing the posterior density of a particular object pose with a convolutional neural network (CNN) that compares observed and rendered images. The network is trained with the maximum likelihood paradigm. We observe empirically that the CNN does not specialize to the geometry or appearance of specific objects. It can be used with objects of vastly different shapes and appearances, and in different backgrounds. Compared to state-of-the-art, we demonstrate a significant improvement on two different datasets which include a total of eleven objects, cluttered background, and heavy occlusion.
24
Leo Breiman · Machine Learning · 2001 · 126.2K citations · Full text
Equation of State Calculations by Fast Computing Machines
N. Metropolis, Arianna W. Rosenbluth, M. N. Rosenbluth et al. · The Journal of Chemical Physics · 1953 · 37.1K citations
Fully convolutional networks for semantic segmentation
Jonathan Long, Evan Shelhamer, Trevor Darrell · 2015 · 36.9K citations