Hasil untuk "Optics. Light"

M. Miri, A. Alú

Exceptional points in optics Many complex systems operate with loss. Mathematically, these systems can be described as non-Hermitian. A property of such a system is that there can exist certain conditions—exceptional points—where gain and loss can be perfectly balanced and exotic behavior is predicted to occur. Optical systems generally possess gain and loss and so are ideal systems for exploring exceptional point physics. Miri and Alù review the topic of exceptional points in photonics and explore some of the possible exotic behavior that might be expected from engineering such systems. Science, this issue p. eaar7709 BACKGROUND Singularities are critical points for which the behavior of a mathematical model governing a physical system is of a fundamentally different nature compared to the neighboring points. Exceptional points are spectral singularities in the parameter space of a system in which two or more eigenvalues, and their corresponding eigenvectors, simultaneously coalesce. Such degeneracies are peculiar features of nonconservative systems that exchange energy with their surrounding environment. In the past two decades, there has been a growing interest in investigating such nonconservative systems, particularly in connection with the quantum mechanics notions of parity-time symmetry, after the realization that some non-Hermitian Hamiltonians exhibit entirely real spectra. Lately, non-Hermitian systems have raised considerable attention in photonics, given that optical gain and loss can be integrated as nonconservative ingredients to create artificial materials and structures with altogether new optical properties. ADVANCES As we introduce gain and loss in a nanophotonic system, the emergence of exceptional point singularities dramatically alters the overall response, leading to a range of exotic functionalities associated with abrupt phase transitions in the eigenvalue spectrum. Even though such a peculiar effect has been known theoretically for several years, its controllable realization has not been made possible until recently and with advances in exploiting gain and loss in guided-wave photonic systems. As shown in a range of recent theoretical and experimental works, this property creates opportunities for ultrasensitive measurements and for manipulating the modal content of multimode lasers. In addition, adiabatic parametric evolution around exceptional points provides interesting schemes for topological energy transfer and designing mode and polarization converters in photonics. Lately, non-Hermitian degeneracies have also been exploited for the design of laser systems, new nonlinear optics phenomena, and exotic scattering features in open systems. OUTLOOK Thus far, non-Hermitian systems have been largely disregarded owing to the dominance of the Hermitian theories in most areas of physics. Recent advances in the theory of non-Hermitian systems in connection with exceptional point singularities has revolutionized our understanding of such complex systems. In the context of optics and photonics, in particular, this topic is highly important because of the ubiquity of nonconservative elements of gain and loss. In this regard, the theoretical developments in the field of non-Hermitian physics have allowed us to revisit some of the well-established platforms with a new angle of utilizing gain and loss as new degrees of freedom, in stark contrast with the traditional approach of avoiding these elements. On the experimental front, progress in fabrication technologies has allowed for harnessing gain and loss in chip-scale photonic systems. These theoretical and experimental developments have put forward new schemes for controlling the functionality of micro- and nanophotonic devices. This is mainly based on the anomalous parameter dependence in the response of non-Hermitian systems when operating around exceptional point singularities. Such effects can have important ramifications in controlling light in new nanophotonic device designs, which are fundamentally based on engineering the interplay of coupling and dissipation and amplification mechanisms in multimode systems. Potential applications of such designs reside in coupled-cavity laser sources with better coherence properties, coupled nonlinear resonators with engineered dispersion, compact polarization and spatial mode converters, and highly efficient reconfigurable diffraction surfaces. In addition, the notion of the exceptional point provides opportunities to take advantage of the inevitable dissipation in environments such as plasmonic and semiconductor materials, which play a key role in optoelectronics. Finally, emerging platforms such as optomechanical cavities provide opportunities to investigate exceptional points and their associated phenomena in multiphysics systems. Ubiquity of non-Hermitian systems, supporting exceptional points, in photonics. (A) A generic non-Hermitian optical system involving two coupled modes with different detuning, ±ω1,2, and gain-loss values, ±γ1,2, coupled at rate of μ. The real part of the associated eigenvalues in a two-dimensional parameter space of the system, revealing the emergence of an exceptional point (EP) singularity. a1 and a2 are the modal amplitudes. (B to E) A range of different photonic systems, which are all governed by the coupled-mode equations. (B) Two coupled lasers pumped at different rates. (C) Dynamical interaction between optical and mechanical degrees of freedom in an optomechanical cavity. (D) A resonator with counter-rotating whispering gallery modes. CW, clockwise; CCW, counterclockwise. (E) A thin metasurface composed of coupled nanoantennas as building blocks. CREDITS: IMAGE IN (A) BASED ON A CONCEPT FROM H. HODAEI ET AL., SCIENCE 346, 975 (2014); IMAGE IN (D) BASED ON CONCEPTS FROM W. CHEN ET AL., NATURE 548, 192 (2017). Exceptional points are branch point singularities in the parameter space of a system at which two or more eigenvalues, and their corresponding eigenvectors, coalesce and become degenerate. Such peculiar degeneracies are distinct features of non-Hermitian systems, which do not obey conservation laws because they exchange energy with the surrounding environment. Non-Hermiticity has been of great interest in recent years, particularly in connection with the quantum mechanical notion of parity-time symmetry, after the realization that Hamiltonians satisfying this special symmetry can exhibit entirely real spectra. These concepts have become of particular interest in photonics because optical gain and loss can be integrated and controlled with high resolution in nanoscale structures, realizing an ideal playground for non-Hermitian physics, parity-time symmetry, and exceptional points. As we control dissipation and amplification in a nanophotonic system, the emergence of exceptional point singularities dramatically alters their overall response, leading to a range of exotic optical functionalities associated with abrupt phase transitions in the eigenvalue spectrum. These concepts enable ultrasensitive measurements, superior manipulation of the modal content of multimode lasers, and adiabatic control of topological energy transfer for mode and polarization conversion. Non-Hermitian degeneracies have also been exploited in exotic laser systems, new nonlinear optics schemes, and exotic scattering features in open systems. Here we review the opportunities offered by exceptional point physics in photonics, discuss recent developments in theoretical and experimental research based on photonic exceptional points, and examine future opportunities in this area from basic science to applied technology.

A. Shaltout, V. Shalaev, M. Brongersma

Dynamic metasurfaces Optical metasurfaces have opened an entirely new field in the quest to manipulate light. Optical metasurfaces can locally impart changes to the amplitude, phase, and polarization of propagating waves. To date, most of these metasurfaces have been passive, with the optical properties largely set in the fabrication process. Shaltout et al. review recent developments toward time-varying metasurfaces and explore the opportunities that adding dynamic control can offer in terms of actively controlling the flow of light. Science, this issue p. eaat3100 BACKGROUND Metasurfaces have opened up a number of remarkable new approaches to manipulate light. These flat optical elements are constructed from a dense array of strongly scattering metallic or semiconductor nanostructures that can impart local changes to the amplitude, phase, and polarization state of light waves. They have facilitated a relaxation of the fundamental Snell’s law for light refraction and enabled the creation of small form factor optical systems capable of performing many tasks that currently can only be achieved with bulky optical components. New photon management capabilities have also emerged, including the achievement of multiple optical functions within a single metasurface element, the realization of very high-numerical apertures, and dispersion engineering with metasurface building blocks. Despite this impressive progress, most metasurfaces we see today are static in nature, and their optical properties are set in stone during their fabrication. However, we are currently witnessing an evolution from passive metasurfaces to active metasurface devices. This natural progression stems from the notion that space and time play complementary roles in Maxwell’s equations. It suggests that structuring materials in both space and time can bring forth new physical phenomena and further broaden the range of possible applications. This Review discusses what is required to create high-performance spatiotemporal metasurfaces and analyzes what new applications and physics they have to offer. ADVANCES To realize the dream of dynamically controlled metasurfaces, we need to achieve strong and tunable light-matter interactions in ultrathin layers of material. In doing so, we cannot rely on the long interactions of lengths and times provided by bulk optical crystals or waveguides. This has stimulated much research aimed at identifying new materials and nanostructures capable of providing dramatically enhanced light-matter interaction and highly tunable optical responses. There are already well-established ways to boost light-matter interaction through the engineering of plasmonic and Mie-style resonances in metallic and semiconductor nanostructures. However, the best approaches to dynamically alter their optical response are the topic of current study. We highlight different approaches that involve electrical gating, optical pumping, mechanical actuation, stimulating phase transitions, magneto-optical effects, electrochemical metallization, liquid-crystal control, and nanostructured nonlinearities. We also discuss how metasurfaces can be used to realize reconfigurable devices, such as tunable lenses and holograms, optical phase modulators, and polarization converters. In addition to these emerging applications, it has become apparent that temporal control of metasurfaces at an ultrafast speed can unlock entirely new physical effects that are not accessible in their static counterparts. Photons interacting with spatiotemporally modulated metasurfaces can display changes in their frequency as well as their linear momentum, angular momentum, and spin. This opens the door to new operating regimes for metasurfaces in which light can experience Doppler shifts, break Lorentz reciprocity, or produce time-reversed optical beams. OUTLOOK The emergence of active metasurfaces is very timely given the many applications that would benefit from having tunable optical components that are flat and easy to integrate. These include a variety of wearables, autonomous vehicles, robotics, augmented and virtual reality, sensing, imaging, and display technologies. However, a massive challenge lies ahead toward realizing the full technological potential of these new elements. The ultimate unit cell of an active metasurface should be subwavelength in size and facilitate large, dynamic amplitude and phase tuning. For larger metasurfaces, the need to individually address and activate the massive number of tiny unit cells will also pose integration and power-consumption challenges that rival those that are currently faced by the semiconductor industry in the creation of the next generation of integrated circuits. If realized, such elements may radically outperform conventional systems that are based on bulky optical and mechanical parts. As new physical effects appear in spatiotemporal metasurfaces, new fundamental questions are also bound to arise. It is already clear that the very basic processes of light absorption, modulation, fluorescent and thermal emission, frequency conversion, and polarization conversion can be manipulated in new ways. As a result, the typically assumed limits for time-invariant or reciprocal systems will need to be reconsidered in these dynamic systems. A concerted, highly interdisciplinary effort is thus required to uncover and push the bounds by which these sheets of spatiotemporally structured materials can manipulate light. Optical phenomena that can be realized with spatiotemporal metasurfaces. Wavelength conversion emulating a Doppler shift, nonreciprocal transmission, and active steering of optical beams. Optical metasurfaces have provided us with extraordinary ways to control light by spatially structuring materials. The space-time duality in Maxwell’s equations suggests that additional structuring of metasurfaces in the time domain can even further expand their impact on the field of optics. Advances toward this goal critically rely on the development of new materials and nanostructures that exhibit very large and fast changes in their optical properties in response to external stimuli. New physics is also emerging as ultrafast tuning of metasurfaces is becoming possible, including wavelength shifts that emulate the Doppler effect, Lorentz nonreciprocity, time-reversed optical behavior, and negative refraction. The large-scale manufacturing of dynamic flat optics has the potential to revolutionize many emerging technologies that require active wavefront shaping with lightweight, compact, and power-efficient components.

M. Khorasaninejad, F. Capasso

C. Rüter, K. Makris, R. El-Ganainy et al.

H. Bisoyi, Quan Li

L. Hau, S. Harris, Z. Dutton et al.

G. Mourou, T. Tajima, S. V. Bulanov

J. Dalibard, Y. Castin, K. Mølmer

P. McMahon

Optical computing has the potential to be faster and more energy-efficient than conventional digital-electronic computing for certain applications. This Perspective article surveys the differences between optics and electronics that could be exploited, and explores the physics and engineering challenges in realizing useful optical computers. There has been a resurgence of interest in optical computing since the early 2010s, both in academia and in industry, with much of the excitement centred around special-purpose optical computers for neural-network processing. Optical computing has been a topic of periodic study since the 1960s, including for neural networks in the 1980s and early 1990s, and a wide variety of optical-computing schemes and architectures have been proposed. In this Perspective article, we provide a systematic explanation of why and how optics might be able to give speed or energy-efficiency benefits over electronics for computing, enumerating 11 features of optics that can be harnessed when designing an optical computer. One often-mentioned motivation for optical computing — that the speed of light is fast — is emphatically not a key differentiating physical property of optics for computing; understanding where an advantage could come from is more subtle. We discuss how gaining an advantage over state-of-the-art electronic processors will likely only be achievable by careful design that harnesses more than 1 of the 11 features, while avoiding a number of pitfalls that we describe.

Peng Wang, Yuanlin Zheng, Xianfeng Chen et al.

Moiré lattices consist of two superimposed identical periodic structures with a relative rotation angle. Moiré lattices have several applications in everyday life, including artistic design, the textile industry, architecture, image processing, metrology and interferometry. For scientific studies, they have been produced using coupled graphene–hexagonal boron nitride monolayers1,2, graphene–graphene layers3,4 and graphene quasicrystals on a silicon carbide surface5. The recent surge of interest in moiré lattices arises from the possibility of exploring many salient physical phenomena in such systems; examples include commensurable–incommensurable transitions and topological defects2, the emergence of insulating states owing to band flattening3,6, unconventional superconductivity4 controlled by the rotation angle7,8, the quantum Hall effect9, the realization of non-Abelian gauge potentials10 and the appearance of quasicrystals at special rotation angles11. A fundamental question that remains unexplored concerns the evolution of waves in the potentials defined by moiré lattices. Here we experimentally create two-dimensional photonic moiré lattices, which—unlike their material counterparts—have readily controllable parameters and symmetry, allowing us to explore transitions between structures with fundamentally different geometries (periodic, general aperiodic and quasicrystal). We observe localization of light in deterministic linear lattices that is based on flat-band physics6, in contrast to previous schemes based on light diffusion in optical quasicrystals12, where disorder is required13 for the onset of Anderson localization14 (that is, wave localization in random media). Using commensurable and incommensurable moiré patterns, we experimentally demonstrate the two-dimensional localization–delocalization transition of light. Moiré lattices may feature an almost arbitrary geometry that is consistent with the crystallographic symmetry groups of the sublattices, and therefore afford a powerful tool for controlling the properties of light patterns and exploring the physics of periodic–aperiodic phase transitions and two-dimensional wavepacket phenomena relevant to several areas of science, including optics, acoustics, condensed matter and atomic physics. A superposition of tunable photonic lattices is used to create optical moiré patterns and demonstrate the resulting localization of light waves through a mechanism based on flat-band physics.

P. Moreau, E. Toninelli, T. Gregory et al.

The production of pairs of entangled photons simply by focusing a laser beam onto a crystal with a nonlinear optical response was used to test quantum mechanics and to open new approaches in imaging. The development of the latter was enabled by the emergence of single-photon-sensitive cameras that are able to characterize spatial correlations and high-dimensional entanglement. Thereby, new techniques emerged, such as ghost imaging of objects — in which the quantum correlations between photons reveal the image from photons that have never interacted with the object — or imaging with undetected photons by using nonlinear interferometers. In addition, quantum approaches in imaging can also lead to an improvement in the performance of conventional imaging systems. These improvements can be obtained by means of image contrast, resolution enhancement that exceeds the classical limit and acquisition of sub-shot-noise phase or amplitude images. In this Review, we discuss the application of quantum states of light for advanced imaging techniques.Using quantum states of light for imaging both reveals quantum phenomena and enables new protocols that result in images that surpass classical limitations. Such systems require both quantum light sources and often the ingenious use of detector technologies.Key pointsImprovements in available camera technologies have enabled the efficient detection and characterization of quantum behaviours in continuous spatial variables.The use of cameras in the context of quantum optics allows the detection and use of high-dimensional quantum states.Quantum states of light can be harnessed to implement quantum imaging protocols that allow improved imaging over classical techniques; such protocols can lead to improved estimation of the transmission, reflectance and phase of an imaged object, in addition to offering improved resolution images of the object.Quantum imaging techniques allow new types of imaging, such as ghost imaging, quantum imaging with undetected photons or the implementation of interaction-free measurements in the context of imaging.Sources of pairs of photons with different wavelengths allow the lack of high-fidelity detectors at exotic wavelengths to be overcome through ghost imaging techniques and quantum nonlinear interferometric imaging techniques.

R. Tyson

History and Background: Introduction. History. Physical Optics: Propagation with Aberrations. Imaging with Aberrations. Representing the Wavefront. Interference. Adaptive Optics Terms. Sources of Aberrations: Atmospheric Turbulence: Descriptions of Atmospheric Turbulence. Refractive Index Structure Constant. Turbulence Effects. Turbulence MTF. Thermal Blooming: Blooming Strength and Critical Power. Turbulence, Jitter, and Thermal Blooming. Non-atmospheric Sources: Optical Misalignments and Jitter. Thermally Induced Distortions of Optics. Manufacturing and Microerrors. Other Sources of Aberrations. Adaptive Optics Compensation: Phase Conjugation. Limitations of Phase Conjugation: Turbulence Spatial Error. Turbulence Temporal Error. Sensor Noise Limitations. Thermal Blooming Compensation. Artificial Guide Stars. Combining the Limitations. Linear Analysis of Random Wavefronts. Linear Analysis of Deterministic Wavefronts: Partial Phase Conjugation. Adaptive Optics Systems: Adaptive Optics Imaging Systems. Beam Propagation Systems: Local Loop Beam Cleanup Systems. Alternative Concepts. Pros and Cons of the Various Approaches. Unconventional Adaptive Optics: Nonlinear Optics. Elastic Photon Scattering, DFWM. Inelastic Photon Scattering. System Engineering. Wavefront Sensing: Directly Measuring Phase: The Non-uniqueness of the Diffraction Pattern. Determining Phase Information from Intensity. Modal and Zonal Sensing. Direct Wavefront Sensing--Modal: Importance of Wavefront Tilt. Measurement of Tilt. Focus Sensing. Modal Sensing of Higher-Order Aberrations. Direct Wavefront Sensing--Zonal: Interferometiric Wavefront Sensing. Hartman Wavefront Sensors. Curvature Sensing. Selecting a Method. Indirect Wavefront Sensing Methods: Multidither Adaptive Optics. Image Sharpening. Wavefront Sampling: Beamsplitters. Hole Gratings. Temporal Duplexing. Reflective Wedges. Diffraction Gratings. Hybrids. Sensitivities of Sampler Concepts. Detectors and Noise. WavefrontCorrection: Modal Tilt Correction. Modal Higher-Order Correction. Segmented Mirrors. Deformable Mirrors: Actuation Techniques. Actuator Influence Functions. Bimorph Corrector Mirrors. Membrane and Micromachine Mirrors. Edge Actuated Mirrors. Large Correcting Optics. Special Correction Devices: Liquid Crystal Phase Modulators. Spatial Light Modulators. Charged-large-array-mirrors. Reconstruction and Controls: Introduction. Single-Channel Linear Control: Fundamental Control Tools. Transfer Functions. Proportional Control. First- and Second-Order Lag. Feedback. Frequency Response of Control Systems. Digital Controls. Multivariate Adaptive Optics Controls: Solution of Linear Equations. Direct Wavefront Reconstruction: Phase from Wavefront Slopes. Modes from Wavefront Slopes. Phase from Wavefront Modes. Modes from Wavefront Modes. Zonal Corrector from Continuous Phase. Modal Corrector from Continuous Phase. Zonal Corrector from Modal Phase. Modal Correctors from Modal Phase. Indirect Reconstructions.Modal Corrector from Wavefront Modes. Zonal Corrector from Wavefront Slopes. Spatiotemporal Considerations. Subject Index.

Tzu-Chi Huang, Wei-Lun Wei, Sunny Saurabh et al.

We investigate the exotic emission behaviors of gadolinium gallium garnet (Gd3Ga5O12, GGG) wafer via X-ray absorption near edge structure (XANES) and X-ray excited optical luminescence (XEOL) in an X-ray nanoprobe beamline. The XANES spectra of the Gd L3 edge and Ga K-edge confirm that both Gd and Ga ions are in the trivalent state in the GGG (100) wafer. The XEOL spectrum exhibits numerous zero-phonon lines (ZPLs) caused by Gd3+ ion transitions. Notably, we observe the quantum cutting phenomenon of Gd3+ ion transitions resulting in red emission due to the 6GJ → 6PJ transition and ultraviolet photon emission due to the 6PJ → 8S7/2 transition. Furthermore, the emission intensities of 6GJ → 6PJ and 6PJ → 8S7/2 transitions increased rapidly with X-ray irradiation time, as observed with time-dependent XEOL. The advantages of using an X-ray nanoprobe, including high peak power density and excellent spatial resolution, allow us to construct XEOL maps to position the enhancing emission intensities of ZPLs in the local area of interest, which can be tuned via X-ray irradiation time. These peculiar emission behaviors of ZPLs in Gd3+ ion transitions may open new avenues in leveraging X-ray nanoprobes to explore quantum materials.

I. Afek, O. Ambar, Y. Silberberg

Xiaochun Liu, Ibón Guillén, Marco La Manna et al.

Non-line-of-sight imaging allows objects to be observed when partially or fully occluded from direct view, by analysing indirect diffuse reflections off a secondary relay surface. Despite many potential applications1–9, existing methods lack practical usability because of limitations including the assumption of single scattering only, ideal diffuse reflectance and lack of occlusions within the hidden scene. By contrast, line-of-sight imaging systems do not impose any assumptions about the imaged scene, despite relying on the mathematically simple processes of linear diffractive wave propagation. Here we show that the problem of non-line-of-sight imaging can also be formulated as one of diffractive wave propagation, by introducing a virtual wave field that we term the phasor field. Non-line-of-sight scenes can be imaged from raw time-of-flight data by applying the mathematical operators that model wave propagation in a conventional line-of-sight imaging system. Our method yields a new class of imaging algorithms that mimic the capabilities of line-of-sight cameras. To demonstrate our technique, we derive three imaging algorithms, modelled after three different line-of-sight systems. These algorithms rely on solving a wave diffraction integral, namely the Rayleigh–Sommerfeld diffraction integral. Fast solutions to Rayleigh–Sommerfeld diffraction and its approximations are readily available, benefiting our method. We demonstrate non-line-of-sight imaging of complex scenes with strong multiple scattering and ambient light, arbitrary materials, large depth range and occlusions. Our method handles these challenging cases without explicitly inverting a light-transport model. We believe that our approach will help to unlock the potential of non-line-of-sight imaging and promote the development of relevant applications not restricted to laboratory conditions. Algorithms based on diffractive wave propagation of light offer effective imaging of complex scenes hidden from direct view.

A. Faraon, I. Fushman, D. Englund et al.

Quantum dots in photonic crystals are interesting because of their potential in quantum information processing and as a testbed for cavity quantum electrodynamics. Recent advances in controlling and coherent probing of such systems open the possibility of realizing quantum networks originally proposed for atomic systems. Here, we demonstrate that non-classical states of light can be coherently generated using a quantum dot strongly coupled to a photonic crystal resonator. We show that the capture of a single photon into the cavity affects the probability that a second photon is admitted. This probability drops when the probe is positioned at one of the two energy eigenstates corresponding to the vacuum Rabi splitting, a phenomenon known as photon blockade, the signature of which is photon antibunching. In addition, we show that when the probe is positioned between the two eigenstates, the probability of admitting subsequent photons increases, resulting in photon bunching. We call this process photon-induced tunnelling. This system represents an ultimate limit for solid-state nonlinear optics at the single-photon level. Along with demonstrating the generation of non-classical photon states, we propose an implementation of a single-photon transistor in this system.

M. Martínez-Corral, B. Javidi

Norasmahan Muridan, Abdul Hadi Sulaiman, Siti Fatimah Norizan et al.

We proposed the generation of a tunable channel-spacing in a multiwavelength fiber laser that incorporates a semiconductor optical amplifier (SOA) and a parallel Lyot filter. Previously, only a few works demonstrated channel spacing tunability using parallel Lyot filter, with none of them utilizing SOA. A stable and tunable multiwavelength spectrum with up to three distinct channel spacings is demonstrated using three different sets of parallel Lyot filter either Short, Long, and Mixed based on varying lengths of polarization-maintaining fiber (PMF). Channel spacing tunability is achieved by selecting different PMF length combinations. Experimental results show that two channel spacing modes, either single or multiple, can be selected for each configuration. Additionally, increasing the SOA drive current results in a greater number of lasing lines with higher intensity within the cavity. The system demonstrates good stability, with peak power differences of 1.46 dB, 0.65 dB, and 2.61 dB for the Short, Long, and Mixed sets, respectively, during a 60-minute observation period.

Ethan Tseng, Ali Mosleh, Fahim Mannan et al.

Most modern commodity imaging systems we use directly for photography—or indirectly rely on for downstream applications—employ optical systems of multiple lenses that must balance deviations from perfect optics, manufacturing constraints, tolerances, cost, and footprint. Although optical designs often have complex interactions with downstream image processing or analysis tasks, today’s compound optics are designed in isolation from these interactions. Existing optical design tools aim to minimize optical aberrations, such as deviations from Gauss’ linear model of optics, instead of application-specific losses, precluding joint optimization with hardware image signal processing (ISP) and highly parameterized neural network processing. In this article, we propose an optimization method for compound optics that lifts these limitations. We optimize entire lens systems jointly with hardware and software image processing pipelines, downstream neural network processing, and application-specific end-to-end losses. To this end, we propose a learned, differentiable forward model for compound optics and an alternating proximal optimization method that handles function compositions with highly varying parameter dimensions for optics, hardware ISP, and neural nets. Our method integrates seamlessly atop existing optical design tools, such as Zemax. We can thus assess our method across many camera system designs and end-to-end applications. We validate our approach in an automotive camera optics setting—together with hardware ISP post processing and detection—outperforming classical optics designs for automotive object detection and traffic light state detection. For human viewing tasks, we optimize optics and processing pipelines for dynamic outdoor scenarios and dynamic low-light imaging. We outperform existing compartmentalized design or fine-tuning methods qualitatively and quantitatively, across all domain-specific applications tested.

Dieter Bimberg, Fumio Koyama, Kenichi Iga