Improve Noise Tolerance of Robust Loss via Noise-Awareness

Understand

Robust loss minimization is an important strategy for handling robust learning issue on noisy labels.

• Current approaches for designing robust losses involve the introduction of noise-robust factors, i.e., hyperparameters, to control the trade-off between noise robustness and learnability.
• However, finding suitable hyperparameters for different datasets with noisy labels is a challenging and time-consuming task.
• Moreover, existing robust loss methods usually assume that all training samples share common hyperparameters, which are independent of instances.