Back to results
US 12,718,388 B2Granted
Non-line-of-sight imaging via neural transient field
AssigneeShanghaitech University
InventorsSiyuan Shen, Zi Wang, Shiying Li, Jingyi Yu
Life of this patent
Estimating the term…
Abstract
A novel neural modeling framework Neural Transient Field (NeTF) is provided for non-line-of-sight (NLOS) imaging. NeTF recovers the 5D transient function in both spatial location and direction, and the training data input is parametrized on the spherical wave-fronts. A Markov chain Monte Carlo (MCMC) algorithm is used to account for sparse and unbalanced sampling in NeTF.
Claims18
A computer-implemented method for imaging a non-line-of-sight (NLOS) scene, comprising:
- encoding, by a computing system, a neural transient field onto a Multi-Layer Perceptron (MLP), wherein the neural transient field represents the NLOS scene as a continuous 5D function of transient;
- feeding a plurality of transient pixels captured by a time-resolved detector from a plurality of detection spots on a relay wall to the MLP;
- outputting a volume density and a surface reflectance along a direction by the MLP in accordance with the plurality of transient pixels;
- and
- reconstructing the NLOS scene in accordance with the volume density and the surface reflectance.
2 more elements
Dependencies are inferred from the claim text.