- Keywords: controlled experiments, robust deep learning, corrupted label, real-world noisy data
- TL;DR: We establish a benchmark of controlled real noise and reveal several interesting findings about real-world noisy data.
- Abstract: Performing controlled experiments on noisy data is essential in thoroughly understanding deep learning across a spectrum of noise levels. Due to the lack of suitable datasets, previous research have only examined deep learning on controlled synthetic noise, and real-world noise has never been systematically studied in a controlled setting. To this end, this paper establishes a benchmark of real-world noisy labels at 10 controlled noise levels. As real-world noise possesses unique properties, to understand the difference, we conduct a large-scale study across a variety of noise levels and types, architectures, methods, and training settings. Our study shows that: (1) Deep Neural Networks (DNNs) generalize much better on real-world noise. (2) DNNs may not learn patterns first on real-world noisy data. (3) When networks are fine-tuned, ImageNet architectures generalize well on noisy data. (4) Real-world noise appears to be less harmful, yet it is more difficult for robust DNN methods to improve. (5) Robust learning methods that work well on synthetic noise may not work as well on real-world noise, and vice versa. We hope our benchmark, as well as our findings, will facilitate deep learning research on noisy data.