The manipulation of the polarization of light is crucial for the development and enhancement of various optical applications. However, existing polarization modulation devices face limitations when it comes to manipulating the spatial distribution of the polarization state of an optical field. Additionally, these devices are only effective for predefined input polarization states, making them inadequate for unpredictable spatially varying polarization fields. Recognizing these challenges, a research team at UCLA has recently presented a groundbreaking design for a polarization transformer diffractive network, offering a new route for universal polarization transformations of spatially varying polarization fields.
The diffractive polarization transformer developed by the UCLA researchers utilizes a series of isotropic diffractive layers, each containing thousands of optimizable transmission coefficients and diffractive features at the subwavelength level. These layers are trained through supervised deep learning. The diffractive processor volume also includes 2D arrays of linear polarizers at different angles to introduce polarization anisotropy. This unique optical architecture allows for the synthesis of a large set of complex-valued, arbitrarily-selected polarization scattering matrices within a compact volume.
To validate the feasibility of the diffractive polarization transformer, the UCLA researchers conducted experiments using wire-grid polarizers fabricated through photolithography and 3D-printed diffractive layers. The proof-of-concept design successfully performed a user-defined polarization permutation operation at the terahertz part of the electromagnetic spectrum. The measured output fields matched the numerical simulations and design objectives, demonstrating the effectiveness of the polarization transformations achieved through this diffractive network.
Moving forward, the UCLA research team plans to improve their designs to operate under broadband illumination, enabling simultaneous processing of the amplitude, phase, polarization, and spectral features encoded in optical fields. By enhancing the diffractive polarization transformer, they aim to develop intelligent machine vision systems with polarization-aware object detection and classification features. These advancements could have significant applications in remote sensing, security/defense, material inspection, and medical imaging.
The development of the diffractive polarization transformer at UCLA opens up new possibilities for the universal transformation of spatially varying polarization fields. By utilizing a unique optical architecture and deep-learning-based training, this technology overcomes the limitations of existing polarization modulation devices and allows for the synthesis of a wide range of polarization scattering matrices. The successful experimental validation indicates the promising potential of this innovation for advancing optical technologies. As the UCLA research team continues to refine their designs and explore broader applications, the future of polarization manipulation in optics looks brighter than ever.
Leave a Reply