Instance-Dependent Noisy Label Learning via Graphical Modelling
This addresses the problem of noisy labels in computer vision, particularly instance-dependent noise, which is critical due to its dependence on image information, but the approach is incremental as it builds on existing noisy-label learning methods.
The paper tackles instance-dependent noisy labels in deep learning by proposing InstanceGM, a graphical modeling approach that combines discriminative and generative models, achieving better accuracy than competitors in most experiments on synthetic and real-world benchmarks.
Noisy labels are unavoidable yet troublesome in the ecosystem of deep learning because models can easily overfit them. There are many types of label noise, such as symmetric, asymmetric and instance-dependent noise (IDN), with IDN being the only type that depends on image information. Such dependence on image information makes IDN a critical type of label noise to study, given that labelling mistakes are caused in large part by insufficient or ambiguous information about the visual classes present in images. Aiming to provide an effective technique to address IDN, we present a new graphical modelling approach called InstanceGM, that combines discriminative and generative models. The main contributions of InstanceGM are: i) the use of the continuous Bernoulli distribution to train the generative model, offering significant training advantages, and ii) the exploration of a state-of-the-art noisy-label discriminative classifier to generate clean labels from instance-dependent noisy-label samples. InstanceGM is competitive with current noisy-label learning approaches, particularly in IDN benchmarks using synthetic and real-world datasets, where our method shows better accuracy than the competitors in most experiments.