Optical micro-phase-shift dropvolume in a diffractive deep neural network.

To provide a desirable number of parallel subnetworks as required to reach a robust inference in an active modulation diffractive deep neural network, a random micro-phase-shift dropvolume that involves five-layer statistically independent dropconnect arrays is monolithically embedded into the unitary backpropagation, which does not require any mathematical derivations with respect to the multilayer arbitrary phase-only modulation masks, even maintaining the nonlinear nested characteristic of neural networks, and generating an opportunity to realize a structured-phase encoding within the dropvolume. Further, a drop-block strategy is introduced into the structured-phase patterns designed to flexibly configure a credible macro-micro phase dropvolume allowing for convergence. Concretely, macro-phase dropconnects concerning fringe griddles that encapsulate sparse micro-phase are implemented. We numerically validate that macro-micro phase encoding is a good plan to the types of encoding within a dropvolume.

Optical random micro-phase-shift DropConnect in a diffractive deep neural network.

The formulation and training of unitary neural networks is the basis of an active modulation diffractive deep neural network. In this Letter, an optical random phase DropConnect is implemented on an optical weight to manipulate a jillion of optical connections in the form of massively parallel sub-networks, in which a micro-phase assumed as an essential ingredient is drilled into Bernoulli holes to enable training convergence, and malposed deflections of the geometrical phase ray are reformulated constantly in epochs, allowing for enhancement of statistical inference. Optically, the random micro-phase-shift acts like a random phase sparse griddle with respect to values and positions, and is operated in the optical path of a projective imaging system. We investigate the performance of the full-drilling and part-drilling phenomena. In general, random micro-phase-shift part-drilling outperforms its full-drilling counterpart both in the training and inference since there are more possible recombinations of geometrical ray deflections induced by random phase DropConnect.

Optical random phase dropout in a diffractive deep neural network.

Unitary learning is a backpropagation (BP) method that serves to update unitary weights in fully connected deep complex-valued neural networks, meeting a prior unitary in an active modulation diffractive deep neural network. However, the square matrix characteristic of unitary weights in each layer results in its learning belonging to a small-sample training, which produces an almost useless network that has a fairly poor generalization capability. To alleviate such a serious over-fitting problem, in this Letter, optical random phase dropout is formulated and designed. The equivalence between unitary forward and diffractive networks deduces a synthetic mask that is seamlessly compounded with a computational modulation and a random sampling comb called dropout. The dropout is filled with random phases in its zero positions that satisfy the Bernoulli distribution, which could slightly deflect parts of transmitted optical rays in each output end to generate statistical inference networks. The enhancement of generalization benefits from the fact that massively parallel full connection with different optical links is involved in the training. The random phase comb introduced into unitary BP is in the form of conjugation, which indicates the significance of optical BP.

Phase-shift determination for a 4 × 4 intelligent photonic neural network with compatible learning.

Intelligent photonic circuits (IPCs) tuned with an appropriate phase-shift vector could enable a photonic intelligent matrix possibly implemented in multiple neural layers for a task-oriented topologies. A photonic Mach-Zehnder Interferometer (MZI) is a fundamental photonic component in IPCs, whose matrix representation could be broadcasted into an arbitrary matrix that is equipped with an optimized phase-shift vector. The initialized MZIs' phases are tentatively probed between analytical elements and a digital weight matrix that is learned from samples with efficient compatible learning for complex-valued neural networks. Nonlinear least squares is utilized to formulate a phase determination system to refine the optimal phase-shift solutions. The robustness of phase determination system for photonic neural networks is discussed in detail. For a preliminary implementation, a basic 4×4 intelligent photonic neural network is utilized to verify the proof of concept on phase-shift determination in IPC through numerical experiments.

Mucilaginibacter yixingensis sp. nov., isolated from vegetable soil.

A Gram-reaction-negative, aerobic, non-motile, non-spore-forming, rod-shaped bacterium, designated YX-36T, was isolated from a vegetable plot in Yixing, Jiangsu province, China. The strain grew at 15-37 °C (optimally at 37 °C), at pH 6.0-9.5 (optimally at pH 6.5) and in the presence of 0-1% (w/v) NaCl (optimally without NaCl). Phylogenetic analysis based on 16S rRNA gene sequences showed that strain YX-36T was related most closely to Mucilaginibacter herbaticus DR-9T (96.88% similarity), followed by Mucilaginibacter sabulilitoris SMS-12T (95.78%), Mucilaginibacter polysacchareus DR-f3T (95.77%) and Mucilaginibacter polysacchareus DRP28T (95.77%). The DNA G+C content of strain YX-36T was 47.2 mol%. The only isoprenoid quinone was menaquinone 7 (MK-7). The major polar lipids were phosphatidylethanolamine and aminophospholipid. The major fatty acids were iso-C15:0, summed feature 3 (iso-C15:0 2-OH/C16:1ω7c) and iso-C17:0 3-OH. On the basis of phenotypic, chemotaxonomic and phylogenetic data, strain YX-36T represents a novel species of the genus Mucilaginibacter, for which the name Mucilaginibacter yixingensis sp. nov. is proposed. The type strain is YX-36T (=DSM 26809T=CCTCC AB 2012880T).

Robust self-calibration three-dimensional shape measurement in fringe-projection photogrammetry.

Commonly, fringe-projection photogrammetry involves two independent stages: system calibration and measurement. The measurement accuracy largely depends on the calibration procedure. However, the results of system calibration may be unstable in different occasions. In this Letter, we propose a robust self-calibration 3D shape measurement in fringe-projection photogrammetry by combining control and measurement points. The control points with known 3D coordinates are provided on the checkerboard, and the measurement points are identified by absolute phase information in the deformed fringes. The introduction of control points in the nonlinear collinearity equations can be regarded as invariant in the optimization procedure, which enhances the measurement robustness. Compared to the binocular model in fringe-projection technique, moreover, multiple-view ray intersection is utilized to reflect the advantage of photogrammetry in the fringe-projection 3D measurement.

Flexible geometrical calibration for fringe-reflection 3D measurement.

System geometrical calibration is a challenging task in fringe-reflection 3D measurement because the fringe displayed on the LCD screen does not lie within the camera's field of view. Commonly, a flat mirror with markers can accomplish system geometrical calibration. However, the position of the markers must be precisely located by photogrammetry in advance. In this Letter, we introduce a calibration method by use of a markerless flat mirror. Experiments in phase measuring deflectometry demonstrate that the proposed method is simple and flexible.

Three-dimensional shape measurement of aspheric mirrors with fringe reflection photogrammetry.

Three-dimensional (3D) shape measurement of an aspheric mirror with fringe reflection photogrammetry involves three steps: correspondence matching, triangulation, and bundle adjustment. Correspondence matching is realized by absolute phase tracking and triangulation is computed by the intersection of reflection and incidence rays. The main contribution in this paper is constraint bundle adjustment for carefully dealing with lens distortion in the process of ray intersection, as compared to the well-known grating reflection photogrammetry. Additionally, a free frame is proposed to alleviate troublesome system geometrical calibration, and constraint bundle adjustment is operated in the free frame to refine the 3D shape. Simulation and experiment demonstrate that constraint bundle adjustment can improve absolute measurement accuracy of aspheric mirrors.

Fringe inverse videogrammetry based on global pose estimation.

Fringe inverse videogrammetry based on global pose estimation is presented to measure a three-dimensional (3D) coordinate. The main components involve an LCD screen, a tactile probe equipped with a microcamera, and a portable personal computer. The LCD is utilized to display fringes, a microcamera is installed on the tactile probe, and the 3D coordinate of the center of the probe tip can be calculated through the microcamera's pose. Fourier fringe analysis is exploited to complete subpixel location of reference points. A convex-relaxation optimization algorithm is employed to estimate the global camera pose, which guarantees global convergence compared with bundle adjustment, a local pose estimation algorithm. The experiments demonstrate that fringe inverse videogrammetry can measure the 3D coordinate precisely.