A Unified Plug-and-Play Framework for Effective Data Denoising and Robust Abstention
This addresses data noise and uncertainty issues for real-world deployment of DNNs, but it is incremental as it builds on existing filtering and abstention techniques.
The paper tackles the problems of data quality and predictive uncertainty in Deep Neural Networks by proposing a unified filtering framework that leverages data density to denoise training data and abstain from uncertain test predictions, achieving state-of-the-art performance on multiple image classification datasets and CNN architectures.
The success of Deep Neural Networks (DNNs) highly depends on data quality. Moreover, predictive uncertainty makes high performing DNNs risky for real-world deployment. In this paper, we aim to address these two issues by proposing a unified filtering framework leveraging underlying data density, that can effectively denoise training data as well as avoid predicting uncertain test data points. Our proposed framework leverages underlying data distribution to differentiate between noise and clean data samples without requiring any modification to existing DNN architectures or loss functions. Extensive experiments on multiple image classification datasets and multiple CNN architectures demonstrate that our simple yet effective framework can outperform the state-of-the-art techniques in denoising training data and abstaining uncertain test data.