Cognitive relevance transform for population re-targeting

Koporec, Gregor (avtor), Košir, Andrej (avtor), Leonardis, Aleš (avtor), Perš, Janez (avtor)

This work examines the differences between a human and a machine in object recognition tasks. The machine is useful as much as the output classification labels are correct and match the dataset-provided labels. However, very often a discrepancy occurs because the dataset label is different than the one expected by a human. To correct this, the concept of the target user population is introduced. The paper presents a complete methodology for either adapting the output of a pre-trained, state-of-the-art object classification algorithm to the target population or inferring a proper, user-friendly categorization from the target population. The process is called ‘user population re-targeting’. The methodology includes a set of specially designed population tests, which provide crucial data about the categorization that the target population prefers. The transformation between the dataset-bound categorization and the new, population-specific categorization is called the ‘Cognitive Relevance Transform’. The results of the experiments on the well-known datasets have shown that the target population preferred such a transformed categorization by a large margin, that the performance of human observers is probably better than previously thought, and that the outcome of re-targeting may be difficult to predict without actual tests on the target population.

• Jezik: Angleški jezik
• Ključne besede: cognitive relevance, deep learning, crowd-sourcing, target user population, categorization, classification
• Vrsta gradiva: Članek v reviji
• Tipologija: 1.01 - Izvirni znanstveni članek
• Organizacija: FE - Fakulteta za elektrotehniko
• Status publikacije: Objavljeno
• Različica publikacije: Objavljena publikacija
• Datum objave: 19.08.2020
• Leto izida: 2020
• Št. strani: 36 str.
• Številčenje: Vol. 20, iss. 17, art. 4668
• PID: 20.500.12556/RUL-128744
• UDK: 004.8
• ISSN pri članku: 1424-8220
• DOI: 10.3390/s20174668
• COBISS.SI-ID: 38147075
• Datum objave v RUL: 27.07.2021

Gradivo je del revije

• Naslov: Sensors
• Skrajšan naslov: Sensors
• Založnik: MDPI
• ISSN: 1424-8220
• COBISS.SI-ID: 10176278

Licence

• Licenca: CC BY 4.0, Creative Commons Priznanje avtorstva 4.0 Mednarodna
• Povezava: http://creativecommons.org/licenses/by/4.0/deed.sl
• Opis: To je standardna licenca Creative Commons, ki daje uporabnikom največ možnosti za nadaljnjo uporabo dela, pri čemer morajo navesti avtorja.
• Začetek licenciranja: 19.08.2020

Sekundarni jezik

• Jezik: Slovenski jezik
• Ključne besede: kognitivna relevanca, globoko učenje, množično pridobivanje podatkov, populacija ciljnih uporabnikov, kategorizacija, razvrščanje