• Introduction Michelson interferometer is widely used in the modern technology. The interference pattern can give us information about the wavelength of the light. Since the data from the experiments is usually a photo, the image-analysis is very important. The post-processing of the Michelson interferometer data includes two steps, [1] extract the physical parameters from the pattern; [2] generate the interference pattern based on the the given input. In fact, there are some physical phenomena too complicated for people to understand. The machine learning approach provides a physical way for us to understand the world. For the current project, we try to apply CNN in extracting the physical parameters from the images. In addition, we want to apply the CdcGAN in generating the interference pattern. • Challenges The hardest part of the project is to determine the hyperparameters. I have got the constant output from my CNN code. It turned out that the “Dropout” layer and the “Normalization” layer are important in getting rid of the local optimum. In addition, the error is fixed after the training. I have to think how to incorporate the physical laws into the CNN modeling. • Insights: Are there any concrete results you can show at this point? o How is your model performing compared with expectations? For the physical parameter extraction part, I have successfully read the wavelength of the light after training for 200 epochs. However, the relative error is 10%, which is still large for the optical experiments. I hope the error can be smaller than 0.1%. I am not sure whether CNN can be accurate enough. • Plan: Are you on track with your project? o What do you need to dedicate more time to? o What are you thinking of changing, if anything? I haven’t finished the GAN part in generating the image. I always got constant outputs. I haven’t load the experimental image by my current code. My experimental image is not of the idealized shape. I want to change GAN to cdGAN.
Log in or sign up for Devpost to join the conversation.