Effect of Label Noise on Robustness of Deep Neural Network Object Detectors

Bishwo Adhikari, Jukka Peltomäki, Saeed Bakhshi Germi, Esa Rahtu, Heikki Huttunen

Tutkimustuotos: KonferenssiartikkeliScientificvertaisarvioitu

2 Lataukset (Pure)

Abstrakti

Label noise is a primary point of interest for safety concerns in previous works as it affects the robustness of a machine learning system by a considerable amount. This paper studies the sensitivity of object detection loss functions to label noise in bounding box detection tasks. Although label noise has been widely studied in the classification context, less attention is paid to its effect on object detection. We characterize different types of label noise and concentrate on the most common type of annotation error, which is missing labels. We simulate missing labels by deliberately removing bounding boxes at training time and study its effect on different deep learning object detection architectures and their loss functions. Our primary focus is on comparing two particular loss functions: cross-entropy loss and focal loss. We also experiment on the effect of different focal loss hyperparameter values with varying amounts of noise in the datasets and discover that even up to 50% missing labels can be tolerated with an appropriate selection of hyperparameters. The results suggest that focal loss is more sensitive to label noise, but increasing the gamma value can boost its robustness.
AlkuperäiskieliEnglanti
OtsikkoComputer Safety, Reliability, and Security. SAFECOMP 2021 Workshops
AlaotsikkoDECSoS, MAPSOD, DepDevOps, USDAI, and WAISE, York, UK, September 7, 2021, Proceedings
ToimittajatIbrahim Habli, Mark Sujan, Simos Gerasimou, Erwin Schoitsch, Friedemann Bitsch
KustantajaSpringer
Sivut239-250
Sivumäärä12
ISBN (elektroninen)978-3-030-83906-2
ISBN (painettu)978-3-030-83905-5
DOI - pysyväislinkit
TilaJulkaistu - 2021
OKM-julkaisutyyppiA4 Artikkeli konferenssijulkaisussa
TapahtumaInternational Conference on Computer Safety, Reliability, and Security - York, Iso-Britannia
Kesto: 7 syysk. 202110 syysk. 2021

Julkaisusarja

NimiLecture Notes in Computer Science
Vuosikerta12853
ISSN (painettu)0302-9743
ISSN (elektroninen)1611-3349

Conference

ConferenceInternational Conference on Computer Safety, Reliability, and Security
Maa/AlueIso-Britannia
KaupunkiYork
Ajanjakso7/09/2110/09/21

Julkaisufoorumi-taso

  • Jufo-taso 1

Sormenjälki

Sukella tutkimusaiheisiin 'Effect of Label Noise on Robustness of Deep Neural Network Object Detectors'. Ne muodostavat yhdessä ainutlaatuisen sormenjäljen.

Siteeraa tätä