Learning from Noisy Labels with Deep Neural Networks
Year: 2014
Type: preprint
Abstract: The availability of large labeled datasets has allowed Convolutional Network models to achieve impressive recognition results. However, in many settings manual annotation of the data is impractical; instead our data has noisy labels, i.e. there is some freely available label for each image which may or may not be accurate. In this paper, we explore the performance of discriminatively-trained Convnets when trained on such noisy data. We introduce an extra noise layer into the network which adapts the network outputs to match the noisy label distribution. The parameters of this noise layer can be estimated as part of the training process and involve simple modifications to current training infrastructures for deep networks. We demonstrate the approaches on several datasets, including large scale experiments on the ImageNet classification benchmark.
Source: arXiv (Cornell University)
Authors: Sainbayar Sukhbaatar, Rob Fergus
Institutions: New York University, Meta (Israel)
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2395115254
