arXiv subjects

Ori Katz

Publications and source records attributed to Ori Katz.

At least 37 records · Page 2Linked to original sources

arXiv2022

NASDUCK: New Constraints on Axion-like Dark Matter from Floquet Quantum Detector

We report on the first results of the Noble and Alkali Spin Detectors for Ultralight Coherent darK matter (NASDUCK) collaboration. We search for the interactions of Axion-Like Particles (ALPs) with atomic spins using an earth-based precision quantum detector as it traverses through the galactic dark matter halo. The detector is composed of spin-polarized xenon gas which can coherently interact with a background ALP dark matter field and an in-situ rubidium Floquet optical-magnetometer. Conducting a five months-long search, we derive new constraints on ALP-proton and ALP-neutron interactions in the $4\times 10^{-15}-4\times 10^{-12}{~\rm eV/c^2}$ mass range. Our limits on the ALP-proton (ALP-neutron) couplings improve upon previous terrestrial bounds by up to 3 orders of magnitude for masses above $4\times 10^{-14}{~\rm eV/c^2}$ ($4\times 10^{-13}{~\rm eV/c^2}$). Moreover, barring the uncertain supernova constraints, the ALP-proton bound improves on all existing terrestrial and astrophysical limits, partially closing the unexplored region for couplings in the range $10^{-6}~{\rm GeV^{-1}}$ to $2\times 10^{-5}~{\rm GeV^{-1}}$. Finally, we also cast bounds on pseudo-scalar dark matter models in which dark matter is quadratically-coupled to the nucleons.

hep-ph↗

arXiv2022

Spectral Flow on the Manifold of SPD Matrices for Multimodal Data Processing

In this paper, we consider data acquired by multimodal sensors capturing complementary aspects and features of a measured phenomenon. We focus on a scenario in which the measurements share mutual sources of variability but might also be contaminated by other measurement-specific sources such as interferences or noise. Our approach combines manifold learning, which is a class of nonlinear data-driven dimension reduction methods, with the well-known Riemannian geometry of symmetric and positive-definite (SPD) matrices. Manifold learning typically includes the spectral analysis of a kernel built from the measurements. Here, we take a different approach, utilizing the Riemannian geometry of the kernels. In particular, we study the way the spectrum of the kernels changes along geodesic paths on the manifold of SPD matrices. We show that this change enables us, in a purely unsupervised manner, to derive a compact, yet informative, description of the relations between the measurements, in terms of their underlying components. Based on this result, we present new algorithms for extracting the common latent components and for identifying common and measurement-specific components.

cs.LG↗

arXiv2022

Pixel-reassignment in Ultrasound Imaging

We present an adaptation of the pixel-reassignment technique from confocal fluorescent microscopy to coherent ultrasound imaging. The method, Ultrasound Pixel-Reassignment (UPR), provides a resolution and signal to noise (SNR) improvement in ultrasound imaging by computationally reassigning off-focus signals acquired using traditional plane-wave compounding ultrasonography. We theoretically analyze the analogy between the optical and ultrasound implementations of pixel reassignment, and experimentally evaluate the imaging quality on tissue-mimicking acoustic phantoms. We demonstrate that UPR provides a $25\%$ resolution improvement and a $3dB$ SNR improvement in in-vitro scans, without any change in hardware or acquisition scheme.

physics.med-ph↗

arXiv2022

Single-shot diffraction-limited imaging through scattering layers via bispectrum analysis

Recently introduced speckle-correlations based techniques enable noninvasive imaging of objects hidden behind scattering layers. In these techniques the hidden object Fourier amplitude is retrieved from the scattered light autocorrelation, and the lost Fourier phase is recovered via iterative phase-retrieval algorithms, which suffer from convergence to wrong local-minima solutions and cannot solve ambiguities in object-orientation. Here, inspired by notions used in astronomy, we experimentally demonstrate that in addition to Fourier amplitude, the object phase information is naturally and inherently encoded in scattered light bispectrum (the Fourier transform of triple-correlation), and can also be extracted from a single high-resolution speckle pattern, based on which we present a single-shot imaging scheme to deterministically and unambiguously retrieve diffraction-limited images of objects hidden behind scattering layers.

physics.optics↗

arXiv2021

Cold Item Integration in Deep Hybrid Recommenders via Tunable Stochastic Gates

A major challenge in collaborative filtering methods is how to produce recommendations for cold items (items with no ratings), or integrate cold item into an existing catalog. Over the years, a variety of hybrid recommendation models have been proposed to address this problem by utilizing items' metadata and content along with their ratings or usage patterns. In this work, we wish to revisit the cold start problem in order to draw attention to an overlooked challenge: the ability to integrate and balance between (regular) warm items and completely cold items. In this case, two different challenges arise: (1) preserving high quality performance on warm items, while (2) learning to promote cold items to relevant users. First, we show that these two objectives are in fact conflicting, and the balance between them depends on the business needs and the application at hand. Next, we propose a novel hybrid recommendation algorithm that bridges these two conflicting objectives and enables a harmonized balance between preserving high accuracy for warm items while effectively promoting completely cold items. We demonstrate the effectiveness of the proposed algorithm on movies, apps, and articles recommendations, and provide an empirical analysis of the cold-warm trade-off.

cs.IR↗

arXiv2021

Roadmap on Wavefront Shaping and deep imaging in complex media

The last decade has seen the development of a wide set of tools, such as wavefront shaping, computational or fundamental methods, that allow to understand and control light propagation in a complex medium, such as biological tissues or multimode fibers. A vibrant and diverse community is now working on this field, that has revolutionized the prospect of diffraction-limited imaging at depth in tissues. This roadmap highlights several key aspects of this fast developing field, and some of the challenges and opportunities ahead.

physics.optics↗

arXiv2021

GAM: Explainable Visual Similarity and Classification via Gradient Activation Maps

We present Gradient Activation Maps (GAM) - a machinery for explaining predictions made by visual similarity and classification models. By gleaning localized gradient and activation information from multiple network layers, GAM offers improved visual explanations, when compared to existing alternatives. The algorithmic advantages of GAM are explained in detail, and validated empirically, where it is shown that GAM outperforms its alternatives across various tasks and datasets.

cs.CV↗

arXiv2021

Label-free video-rate micro-endoscopy through flexible fibers via Fiber Bundle Distal Holography (FiDHo)

Fiber-based micro-endoscopes are a critically important tool for minimally-invasive deep-tissue imaging. However, the state-of-the-art micro-endoscopes cannot perform three-dimensional imaging through dynamically-bent fibers without the use of bulky optical elements such as lenses and scanners at the distal end, increasing the footprint and tissue-damage. While great efforts have been invested in developing approaches that avoid distal bulky optical elements, the fundamental barrier of dynamic optical wavefront-distortions in propagation through flexible fibers, limits current approaches to nearly-static or non-flexible fibers. Here, we present an approach that allows holographic 3D bend-insensitive, coherence-gated, micro-endoscopic imaging, using commercially available multi-core fibers (MCFs). We achieve this by adding a miniature partially-reflecting mirror to the distal fiber-tip, allowing us to perform low-coherence full-field phase-shifting holography. We demonstrate widefield diffraction-limited reflection imaging of amplitude and phase targets through dynamically bent fibers at video-rates. Our approach holds great potential for label-free investigations of dynamic samples.

physics.optics↗

arXiv2021

Acousto-optic Ptychography

Acousto-optic imaging (AOI) enables optical-contrast imaging deep inside scattering samples via localized ultrasound-modulation of scattered light. While AOI allows optical investigations at depths, its imaging resolution is inherently limited by the ultrasound wavelength, prohibiting microscopic investigations. Here, we propose a novel computational imaging approach that allows to achieve optical diffraction-limited imaging using a conventional AOI system. We achieve this by extracting diffraction-limited imaging information from 'memory-effect' speckle-correlations in the conventionally detected ultrasound-modulated scattered-light fields. Specifically, we identify that since speckle correlations allow to estimate the Fourier-magnitude of the field inside the ultrasound focus, scanning the ultrasound focus enables robust diffraction-limited reconstruction of extended objects using ptychography, i.e. we exploit the ultrasound focus as the scanned spatial-gate 'probe' required for ptychographic phase-retrieval. Moreover, we exploit the short speckle decorrelation-time in dynamic media, which is usually considered a hurdle for wavefront-shaping based approaches, for improved ptychographic reconstruction. We experimentally demonstrate non-invasive imaging of targets that extend well beyond the memory-effect range, with a 40-times resolution improvement over conventional AOI, surpassing the performance of state-of-the-art approaches.

physics.optics↗

arXiv2020

Fourier holographic endoscopy for label-free imaging through a narrow and curved passage

Ultrathin lensless fibre endoscopes offer minimally invasive investigation, but they mostly operate as a rigid type due to the need for prior calibration of a fibre probe. Furthermore, most implementations work in fluorescence mode rather than label-free imaging mode, making them unsuitable for medicine and industry. Herein, we report a fully flexible ultrathin fibre endoscope taking 3D holographic images of unstained tissues with 0.87-μm spatial resolution. Using a bare fibre bundle as thin as 200-μm diameter, we design a lensless Fourier holographic imaging configuration to selectively detect weak reflections from biological tissues, a critical step for stain-free reflectance imaging. A unique algorithm is developed for calibration-free holographic image reconstruction, allowing us to image through a narrow and curved passage regardless of fibre bending. We demonstrate endoscopic reflectance imaging of unstained rat intestine tissues that are completely invisible to conventional endoscopes. The proposed endoscope will expedite more accurate and earlier diagnosis than before with minimal complications.

physics.optics↗

arXiv2020

RecoBERT: A Catalog Language Model for Text-Based Recommendations

Language models that utilize extensive self-supervised pre-training from unlabeled text, have recently shown to significantly advance the state-of-the-art performance in a variety of language understanding tasks. However, it is yet unclear if and how these recent models can be harnessed for conducting text-based recommendations. In this work, we introduce RecoBERT, a BERT-based approach for learning catalog-specialized language models for text-based item recommendations. We suggest novel training and inference procedures for scoring similarities between pairs of items, that don't require item similarity labels. Both the training and the inference techniques were designed to utilize the unlabeled structure of textual catalogs, and minimize the discrepancy between them. By incorporating four scores during inference, RecoBERT can infer text-based item-to-item similarities more accurately than other techniques. In addition, we introduce a new language understanding task for wine recommendations using similarities based on professional wine reviews. As an additional contribution, we publish annotated recommendations dataset crafted by human wine experts. Finally, we evaluate RecoBERT and compare it to various state-of-the-art NLP models on wine and fashion recommendations tasks.

cs.IR↗

arXiv2020

Guidestar-free image-guided wavefront-shaping

Optical imaging through scattering media is a fundamental challenge in many applications. Recently, substantial breakthroughs such as imaging through biological tissues and looking around corners have been obtained by the use of wavefront-shaping approaches. However, these require an implanted guide-star for determining the wavefront correction, controlled coherent illumination, and most often raster scanning of the shaped focus. Alternative novel computational approaches that exploit speckle correlations, avoid guide-stars and wavefront control but are limited to small two-dimensional objects contained within the memory-effect correlations range. Here, we present a new concept, image-guided wavefront-shaping, allowing non-invasive, guidestar-free, widefield, incoherent imaging through highly scattering layers, without illumination control. Most importantly, the wavefront-correction is found even for objects that are larger than the memory-effect range, by blindly optimizing image-quality metrics. We demonstrate imaging of extended objects through highly-scattering layers and multi-core fibers, paving the way for non-invasive imaging in various applications, from microscopy to endoscopy.

physics.optics↗

arXiv2020

Using fiber-bending generated speckles for improved working distance and background-rejection in lensless microendoscopy

Lensless flexible fiber-bundle based endoscopes allow imaging at depths beyond the reach of conventional microscopes with a minimal footprint. These multicore fibers provide a simple solution for widefield fluorescent imaging when the target is adjacent to the fiber facet. However, they suffer from a very limited working distance and out-of-focus background. Here, we carefully study the dynamic speckle illumination patterns generated by bending a commercial fiber-bundle, and show that they can be exploited to allow extended working distance and background rejection, using a super-resolution fluctuations imaging (SOFI) analysis of multiple frames, without the addition of any optical elements.

physics.optics↗

arXiv2020

100,000 frames-per-second compressive imaging with a conventional rolling-shutter camera by random point-spread-function engineering

We demonstrate an approach that allows taking videos at very high-speeds of over 100,000 frames per second (fps) by exploiting the fast sampling rate of the standard rolling-shutter readout mechanism, common to most conventional sensors, and a compressive-sampling acquisition scheme. Our approach is directly applied to a conventional imaging system by the simple addition of a diffuser to the pupil plane, randomly encoding the entire field-of-view to each camera row, while maintaining diffraction-limited resolution. A short video is reconstructed from a single camera frame via a compressed-sensing reconstruction algorithm, exploiting inherent sparsity of the imaged scene.

eess.IV↗

arXiv2020

Attentive Item2Vec: Neural Attentive User Representations

Factorization methods for recommender systems tend to represent users as a single latent vector. However, user behavior and interests may change in the context of the recommendations that are presented to the user. For example, in the case of movie recommendations, it is usually true that earlier user data is less informative than more recent data. However, it is possible that a certain early movie may become suddenly more relevant in the presence of a popular sequel movie. This is just a single example of a variety of possible dynamically altering user interests in the presence of a potential new recommendation. In this work, we present Attentive Item2vec (AI2V) - a novel attentive version of Item2vec (I2V). AI2V employs a context-target attention mechanism in order to learn and capture different characteristics of user historical behavior (context) with respect to a potential recommended item (target). The attentive context-target mechanism enables a final neural attentive user representation. We demonstrate the effectiveness of AI2V on several datasets, where it is shown to outperform other baselines.

cs.IR↗

arXiv2020

Neural Attentive Multiview Machines

An important problem in multiview representation learning is finding the optimal combination of views with respect to the specific task at hand. To this end, we introduce NAM: a Neural Attentive Multiview machine that learns multiview item representations and similarity by employing a novel attention mechanism. NAM harnesses multiple information sources and automatically quantifies their relevancy with respect to a supervised task. Finally, a very practical advantage of NAM is its robustness to the case of dataset with missing views. We demonstrate the effectiveness of NAM for the task of movies and app recommendations. Our evaluations indicate that NAM outperforms single view models as well as alternative multiview methods on item recommendations tasks, including cold-start scenarios.

cs.LG↗

arXiv2019

InverSynth: Deep Estimation of Synthesizer Parameter Configurations from Audio Signals

Sound synthesis is a complex field that requires domain expertise. Manual tuning of synthesizer parameters to match a specific sound can be an exhaustive task, even for experienced sound engineers. In this paper, we introduce InverSynth - an automatic method for synthesizer parameters tuning to match a given input sound. InverSynth is based on strided convolutional neural networks and is capable of inferring the synthesizer parameters configuration from the input spectrogram and even from the raw audio. The effectiveness InverSynth is demonstrated on a subtractive synthesizer with four frequency modulated oscillators, envelope generator and a gater effect. We present extensive quantitative and qualitative results that showcase the superiority InverSynth over several baselines. Furthermore, we show that the network depth is an important factor that contributes to the prediction accuracy.

cs.SD↗

arXiv2019

Scalable Attentive Sentence-Pair Modeling via Distilled Sentence Embedding

Recent state-of-the-art natural language understanding models, such as BERT and XLNet, score a pair of sentences (A and B) using multiple cross-attention operations - a process in which each word in sentence A attends to all words in sentence B and vice versa. As a result, computing the similarity between a query sentence and a set of candidate sentences, requires the propagation of all query-candidate sentence-pairs throughout a stack of cross-attention layers. This exhaustive process becomes computationally prohibitive when the number of candidate sentences is large. In contrast, sentence embedding techniques learn a sentence-to-vector mapping and compute the similarity between the sentence vectors via simple elementary operations. In this paper, we introduce Distilled Sentence Embedding (DSE) - a model that is based on knowledge distillation from cross-attentive models, focusing on sentence-pair tasks. The outline of DSE is as follows: Given a cross-attentive teacher model (e.g. a fine-tuned BERT), we train a sentence embedding based student model to reconstruct the sentence-pair scores obtained by the teacher model. We empirically demonstrate the effectiveness of DSE on five GLUE sentence-pair tasks. DSE significantly outperforms several ELMO variants and other sentence embedding methods, while accelerating computation of the query-candidate sentence-pairs similarities by several orders of magnitude, with an average relative degradation of 4.6% compared to BERT. Furthermore, we show that DSE produces sentence embeddings that reach state-of-the-art performance on universal sentence representation benchmarks. Our code is made publicly available at https://github.com/microsoft/Distilled-Sentence-Embedding.

cs.LG↗