Abstract

Can one hide an averse food in a flavorful food so that the averse food is not perceptible? Here we take a statistical signal processing approach to show how to optimally design a food additive (either using pure flavor compounds or natural ingredients) to act as a steganographic key for this food steganography problem. We use a synthesis-based model of olfaction that has emerged in the psychology literature and the percept known as olfactory white acts as an intermediate signal in our approach. The problem decomposes into predictive analytics and prescriptive analytics components. In the predictive component, we learn a mapping from the space of physic-ochemical descriptors of flavor compounds to the space of perceptual odor descriptors through multivariate regression with nuclear norm regularization. In the prescriptive component, we find optimal mixtures of compounds or foods to make the averse food imperceptible in the flavorful food by posing and solving an inverse problem with non-negativity constraints. We demonstrate the proposed approach on real-world physicochemical and olfactory perception data for compounds in food.

Original languageEnglish (US)
Title of host publication2014 IEEE Workshop on Statistical Signal Processing, SSP 2014
PublisherIEEE Computer Society
Pages21-24
Number of pages4
ISBN (Print)9781479949755
DOIs
StatePublished - 2014
Event2014 IEEE Workshop on Statistical Signal Processing, SSP 2014 - Gold Coast, QLD, Australia
Duration: Jun 29 2014Jul 2 2014

Publication series

NameIEEE Workshop on Statistical Signal Processing Proceedings

Other

Other2014 IEEE Workshop on Statistical Signal Processing, SSP 2014
Country/TerritoryAustralia
CityGold Coast, QLD
Period6/29/147/2/14

Keywords

  • olfactory signal processing
  • steganography

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Applied Mathematics
  • Signal Processing
  • Computer Science Applications

Fingerprint

Dive into the research topics of 'FOOD steganography with olfactory white'. Together they form a unique fingerprint.

Cite this