Towards optimal multimode fiber imaging by leveraging input polarization and deep learning

Deep learning techniques provide a plausible route towards achieving practical imaging through multimode fibers. The results produced by these methods are often influenced by physical factors like temperature, fiber length, external perturbations, and polarization state of the input light. Literature focuses on these different elements impacting deep-learning-enabled multimode imaging, yet the effects of input polarization remain under-explored. Here, we show experimentally that the state of polarization of light, being injected at multimode fiber input, affects the fidelity of reconstructed images from speckle patterns. Certain polarization states produce high-quality images at fiber output, while some yield degraded results. We have designed a conditional generative adversarial network~(CGAN) for image regeneration at various degrees of input light polarization. At a particular polarization state and with a thinner core multimode fiber, our network can reconstruct images with an average structural similarity index(SSIM) exceeding 0.9. Hence, in the case of multimode fibers that are held fixed, optimal imaging can be achieved by leveraging deep learning models with the input light polarization state, where the fidelity of images is maximum. We also show that the model can be trained to image adequately for all input light polarization states when the fiber has bends or twists. We anticipate that our work will be a stepping stone toward developing high-resolution and less invasive multimode fiber endoscopes.