Monocular depth prediction with object removal from single image
Issaranon, Theerasit
Loading…
Permalink
https://hdl.handle.net/2142/105848
Description
Title
Monocular depth prediction with object removal from single image
Author(s)
Issaranon, Theerasit
Issue Date
2019-07-12
Director of Research (if dissertation) or Advisor (if thesis)
Forsyth, David
Department of Study
Electrical & Computer Eng
Discipline
Electrical & Computer Engr
Degree Granting Institution
University of Illinois at Urbana-Champaign
Degree Name
M.S.
Degree Level
Thesis
Keyword(s)
Single-Image Depth Prediction
Object Removal
Occluded Vision
Abstract
We describe a method that predicts, from a single RGB image, a depth map that describes the scene when a masked object is removed – we call this “counterfactual depth” that models hidden scene geometry together with the observations. Our method works for the same reason that scene completion works: the spatial structure of objects is simple. But we offer a much higher resolution representation of space than current scene completion methods, as we operate at pixel-level precision and do not rely on a voxel representation. We can remove objects arbitrarily with an instructed object mask. Furthermore, we do not require RGBD inputs.
Our method uses a standard encoder-decoder architecture, with a decoder modified to accept an object mask. We systematically construct a small evaluation dataset that we have collected. Using this dataset, we show that our depth predictions for masked objects are better than other baselines in the real scene. Given unmasked images, our approach performs comparatively well as a regular scene depth predictor.
Use this login method if you
don't
have an
@illinois.edu
email address.
(Oops, I do have one)
IDEALS migrated to a new platform on June 23, 2022. If you created
your account prior to this date, you will have to reset your password
using the forgot-password link below.
