A Diffractive Neural Network with Weight-Noise-Injection Training
This work addresses noise resistance in optical-based classification for applications like imaging or sensing, but it appears incremental as it builds on existing diffractive neural network methods.
The paper tackles the problem of improving robustness in diffractive neural networks against surface shape errors by proposing Weight Noise Injection training, resulting in a network (SRNN) that maintains higher accuracy under serious noise.
We propose a diffractive neural network with strong robustness based on Weight Noise Injection training, which achieves accurate and fast optical-based classification while diffraction layers have a certain amount of surface shape error. To the best of our knowledge, it is the first time that using injection weight noise during training to reduce the impact of external interference on deep learning inference results. In the proposed method, the diffractive neural network learns the mapping between the input image and the label in Weight Noise Injection mode, making the network's weight insensitive to modest changes, which improve the network's noise resistance at a lower cost. By comparing the accuracy of the network under different noise, it is verified that the proposed network (SRNN) still maintains a higher accuracy under serious noise.