ASIST: Automatic semantically invariant scene transformation

Or Litany; Tal Remez; Daniel Freedman; Lior Shapira; Alex Bronstein; Ran Gal

doi:https://doi.org/10.1016/j.cviu.2016.08.002

ASIST: Automatic semantically invariant scene transformation

Or Litany, Tal Remez, Daniel Freedman, Lior Shapira, Alex Bronstein, Ran Gal

School of Electrical Engineering

Tel Aviv University

Research output: Contribution to journal › Article › peer-review

Abstract

We present ASIST, a technique for transforming point clouds by replacing objects with their semantically equivalent counterparts. Transformations of this kind have applications in virtual reality, repair of fused scans, and robotics. ASIST is based on a unified formulation of semantic labeling and object replacement; both result from minimizing a single objective. We present numerical tools for the efficient solution of this optimization problem. The method is experimentally assessed on new datasets of both synthetic and real point clouds, and is additionally compared to two recent works on object replacement on data from the corresponding papers.

Original language	English
Pages (from-to)	284-299
Number of pages	16
Journal	Computer Vision and Image Understanding
Volume	157
DOIs	https://doi.org/10.1016/j.cviu.2016.08.002
State	Published - 1 Apr 2017

Keywords

Alternating minimization
Iterative closest point
Object recognition
Pose estimation
Random forest
Registration
Semantic segmentation

All Science Journal Classification (ASJC) codes

Software
Signal Processing
Computer Vision and Pattern Recognition

Access to Document

https://doi.org/10.1016/j.cviu.2016.08.002

Cite this

@article{ee544ae3f48f4316bc4cb97ec53c03b7,

title = "ASIST: Automatic semantically invariant scene transformation",

abstract = "We present ASIST, a technique for transforming point clouds by replacing objects with their semantically equivalent counterparts. Transformations of this kind have applications in virtual reality, repair of fused scans, and robotics. ASIST is based on a unified formulation of semantic labeling and object replacement; both result from minimizing a single objective. We present numerical tools for the efficient solution of this optimization problem. The method is experimentally assessed on new datasets of both synthetic and real point clouds, and is additionally compared to two recent works on object replacement on data from the corresponding papers.",

keywords = "Alternating minimization, Iterative closest point, Object recognition, Pose estimation, Random forest, Registration, Semantic segmentation",

author = "Or Litany and Tal Remez and Daniel Freedman and Lior Shapira and Alex Bronstein and Ran Gal",

note = "Publisher Copyright: {\textcopyright} 2016 Elsevier Inc.",

year = "2017",

month = apr,

day = "1",

doi = "https://doi.org/10.1016/j.cviu.2016.08.002",

language = "الإنجليزيّة",

volume = "157",

pages = "284--299",

journal = "Computer Vision and Image Understanding",

issn = "1077-3142",

publisher = "Academic Press Inc.",

}

TY - JOUR

T1 - ASIST

T2 - Automatic semantically invariant scene transformation

AU - Litany, Or

AU - Remez, Tal

AU - Freedman, Daniel

AU - Shapira, Lior

AU - Bronstein, Alex

AU - Gal, Ran

PY - 2017/4/1

Y1 - 2017/4/1

N2 - We present ASIST, a technique for transforming point clouds by replacing objects with their semantically equivalent counterparts. Transformations of this kind have applications in virtual reality, repair of fused scans, and robotics. ASIST is based on a unified formulation of semantic labeling and object replacement; both result from minimizing a single objective. We present numerical tools for the efficient solution of this optimization problem. The method is experimentally assessed on new datasets of both synthetic and real point clouds, and is additionally compared to two recent works on object replacement on data from the corresponding papers.

AB - We present ASIST, a technique for transforming point clouds by replacing objects with their semantically equivalent counterparts. Transformations of this kind have applications in virtual reality, repair of fused scans, and robotics. ASIST is based on a unified formulation of semantic labeling and object replacement; both result from minimizing a single objective. We present numerical tools for the efficient solution of this optimization problem. The method is experimentally assessed on new datasets of both synthetic and real point clouds, and is additionally compared to two recent works on object replacement on data from the corresponding papers.

KW - Alternating minimization

KW - Iterative closest point

KW - Object recognition

KW - Pose estimation

KW - Random forest

KW - Registration

KW - Semantic segmentation

UR - http://www.scopus.com/inward/record.url?scp=84994713319&partnerID=8YFLogxK

U2 - https://doi.org/10.1016/j.cviu.2016.08.002

DO - https://doi.org/10.1016/j.cviu.2016.08.002

M3 - مقالة

SN - 1077-3142

VL - 157

SP - 284

EP - 299

JO - Computer Vision and Image Understanding

JF - Computer Vision and Image Understanding

ER -

ASIST: Automatic semantically invariant scene transformation

Abstract

Keywords

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this