• Publications
  • Influence
Joint Recovery of Dense Correspondence and Cosegmentation in Two Images
TLDR
We propose a new hierarchical Markov random field model to jointly recover cosegmentation and dense per-pixel correspondence in two images. Expand
Neural Inverse Rendering for General Reflectance Photometric Stereo
TLDR
We present a novel convolutional neural network architecture for photometric stereo (Woodham, 1980), a problem of recovering 3D object surface normals from multiple images observed under varying illuminations. Expand
Graph Cut Based Continuous Stereo Matching Using Locally Shared Labels
TLDR
We present an accurate and efficient stereo matching method using locally shared labels, a new labeling scheme that enables spatial propagation in MRF inference using graph cuts. Expand
Continuous 3D Label Stereo Matching Using Local Expansion Moves
TLDR
We present an accurate stereo matching method using <italic>local expansion moves</italic] based on graph cuts. Expand
Fast Multi-frame Stereo Scene Flow with Motion Segmentation
TLDR
We propose a new multi-frame method for efficiently computing scene flow (dense depth and optical flow) and camera ego-motion for a dynamic scene observed from a moving stereo camera rig. Expand
Semi-global Stereo Matching with Surface Orientation Priors
TLDR
We propose a simple extension to the Semi-Global Matching algorithm, SGM-P, that utilizes precomputed surface orientation priors. Expand
Continuous Stereo Matching using Local Expansion Moves
TLDR
We present an accurate and efficient stereo matching method using local expansion moves, a new move making scheme using graph cuts, which produces submodular moves deriving a subproblem optimality. Expand
Path Planning using Neural A* Search
TLDR
In this work, we reformulate a canonical A* search algorithm to be differentiable and couple it with a convolutional encoder to form an end-to-end trainable neural network planner. Expand
Superdifferential cuts for binary energies
TLDR
We propose an efficient and general purpose energy optimization method for binary variable energies used in various low-level vision tasks, and show improvements over state-of-the-art methods. Expand
Neural Photometric Stereo Reconstruction for General Reflectance Surfaces
TLDR
We present a novel convolutional neural network architecture for photometric stereo, a problem of recovering 3D object surface normals from multiple images observed under varying illuminations. Expand
...
1
2
...