Effect of Label Noise on Robustness of Deep Neural Network Object Detectors
Adhikari, Bishwo; Peltomäki, Jukka; Bakhshi Germi, Saeed; Rahtu, Esa; Huttunen, Heikki (2021)
Adhikari, Bishwo
Peltomäki, Jukka
Bakhshi Germi, Saeed
Rahtu, Esa
Huttunen, Heikki
Teoksen toimittaja(t)
Habli, Ibrahim
Sujan, Mark
Gerasimou, Simos
Schoitsch, Erwin
Bitsch, Friedemann
Springer
2021
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-202109066961
https://urn.fi/URN:NBN:fi:tuni-202109066961
Kuvaus
Peer reviewed
Tiivistelmä
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.
Kokoelmat
- TUNICRIS-julkaisut [16740]