Learning to Navigate the Energy Landscape
Authors:
- Julien ValentinJVJulien Valentin
- Angela DaiADAngela Dai
- Matthias NiessnerMNMatthias Niessner
- Pushmeet KohliPKPushmeet Kohli
- Philip TorrPTPhilip Torr
- Shahram IzadiSIShahram Izadi
- Cem KeskinCKCem Keskin
- 1 October 2016
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 323-332
Abstract
In this paper, we present a novel, general, and efficient architecture for addressing computer vision problems that are approached from an 'Analysis by Synthesis' standpoint. Analysis by synthesis involves the minimization of reconstruction error, which is typically a non-convex function of the latent target variables. State-of-the-art methods adopt a hybrid scheme where discriminatively trained predictors like Random Forests or Convolutional Neural Networks are used to initialize local search algorithms. While these hybrid methods have been shown to produce promising results, they often get stuck in local optima. Our method goes beyond the conventional hybrid architecture by not only proposing multiple accurate initial solutions but by also defining a navigational structure over the solution space that can be used for extremely efficient gradient-free local search. We demonstrate the efficacy and generalizability of our approach on tasks as diverse as Hand Pose Estimation, RGB Camera Relocalization, and Image Retrieval.Keywords
All Related Versions
This publication has 40 references indexed in Scilit:
- Optimal Lower Bounds for Locality-Sensitive Hashing (Except When q is Tiny)ACM Transactions on Computation Theory, 2014
- Beyond Locality-Sensitive Hashing2013
- Real-time 3D reconstruction at scale using voxel hashingACM Transactions on Graphics, 2013
- Random Forest for Image Annotation2012
- Simultaneous Segmentation and Pose Estimation of Humans Using Dynamic Graph CutsInternational Journal of Computer Vision, 2008
- Improving the Agility of Keyframe-Based SLAM2008
- Vision as Bayesian inference: analysis by synthesis?Trends in Cognitive Sciences, 2006
- Image segmentation by data-driven markov chain monte carloIEEE Transactions on Pattern Analysis and Machine Intelligence, 2002
- Multidimensional divide-and-conquerCommunications of the ACM, 1980
- An Algorithm for Finding Best Matches in Logarithmic Expected TimeACM Transactions on Mathematical Software, 1977