Abstract:Recently, many benchmark datasets have been found to contain noisy labels caused by unavoidable human mistakes. Many researchers propose new noise-aware loss functions to achieve robust classification performance. However, we discover that existing noise-aware loss functions cannot fully heal the damage caused by the noise. On the other hand, some methods filter out low confidence samples and train new models, whereas the filtered samples contain both noisy and hard samples that are critical for the robustness of models. Based on the above two discoveries, we devised the Noise-aware Network (NA-Net) for robust training with noisy labels. Each layer of NA-Net contains three groups of convolution kernels responsible for mix samples, clean samples, and noisy samples, termed as mix-kernels, clean-kernels, and noise-kernels, respectively. Mix-kernels are used for finding the clean samples with a newly devised noise-immune (NI) loss function; clean-kernels are targeted at learning better features without being misguided by noise; noise-kernels are trained by the remaining samples to rectify wrong labels for the next iteration. Meanwhile, for increasing the classification performance of mix-kernels, the extracted feature maps of clean-kernels without being poisoned are combined as the input of mix-kernels of the next layer. Also, the knowledge distillation strategy is adopted to distill the knowledge from clean-kernels to the noise-kernels. Extensive experiments demonstrate that the mutual promotion of three groups of kernels in NA-Net achieves state-of-the-art performance on both artificial noisy datasets and real-world datasets.

An Algorithm Of Wavelet Network Learning From Noisy Data

A wavelet-based method for de-noising data of electric load modeling

Multiresolution Algorithm of Wavelet Network Learning from Absent Training Data

A novel nonlinear system identification method based on robust wavelet network

Nonlinear System Identification Using Orthogonal Wavelet Networks

Noise is the Fatal Poison: A Noise-aware Network for Noisy Dataset Classification

Orthogonal Wavelet Networks for Nonlinear Dynamic System Identification

WEIGHTED BAND-WISE ORTHOGONAL WAVELET NETWORK BASED ON GRID STRUCTURE FOR NONLINEAR SYSTEMS IDENTIFICATION

New Kind of Convergence Algorithm about Wavelet Neural Network Based on Spectrum Analysis

Construction of a New Adaptive Wavelet Network and Its Learning Algorithm

A NOVEL METHOD FOR IDENTIFYING NONLINEAR SYSTEMS USING ORTHOGONAL WAVELET NETWORKS

A HYBRID LEARNING ALGORITHM BASED ON LEVENBERGMARQUARDT ALGORITHM AND LEAST SQUARES TECHNIQUES FOR WAVELET NETWORKS

Robust Time Series Denoising with Learnable Wavelet Packet Transform

Nonlinear Dynamic System Identification Based on Wavelet Approximation

Identifying nonlinear dynamic system using the improved multiresolution wave-net

Nonlinear Adaptive Noise Suppression Based on Wavelet Transform

Anomaly Detection of Network Traffic Based on Wavelet Packet

Wavelet Integrated CNNs for Noise-Robust Image Classification

M-band Wavelet Network for Machine Anomaly Detection from a Frequency Perspective

A Novel Wavelet Method for Noise Analysis of Nonlinear Circuits

Trainable Wavelet Neural Network for Non-Stationary Signals