The COVID-19 can be transmitted through airborne droplets, aerosols and other carriers. In order to better reduce people's risk of infection, individuals need to wear masks to prevent the spread of the virus when going out to public places, seeking medical treatment, and taking public transportation. This paper is based on the improved RETINAFACE algorithm, which effectively realizes the detection of mask wearing, and on the basis of this algorithm, realizes the function of judging whether the mask is worn correctly. In the face recognition algorithm, this paper designs a face recognition algorithm with higher accuracy. The system combines a face mask wearing detection algorithm, a mask standard wearing detection algorithm and a face recognition algorithm. In addition, this article adds a voice prompt module to better assist the integrity of the system's functions. The test results of the final experiment show that the system can effectively achieve the purpose of face mask detection and recognition.

/pdf/intelligent-detection-and-recognition-system-for-mask-1h3qmtcry4.pdf

Intelligent detection and recognition system for mask wearing based on improved RetinaFace algorithm

Abstract Photonics-based neural networks promise to outperform electronic counterparts, accelerating neural network computations while reducing power consumption and footprint. However, these solutions suffer from physical layer constraints arising from the underlying analog photonic hardware, impacting the resolution of computations (in terms of effective number of bits), requiring the use of positive-valued inputs, and imposing limitations in the fan-in and in the size of convolutional kernels. To abstract these constraints, in this paper we introduce the concept of Photonic-Aware Neural Network (PANN) architectures, i.e., deep neural network models aware of the photonic hardware constraints. Then, we devise PANN training schemes resorting to quantization strategies aimed to obtain the required neural network parameters in the fixed-point domain, compliant with the limited resolution of the underlying hardware. We finally carry out extensive simulations exploiting PANNs in image classification tasks on well-known datasets (MNIST, Fashion-MNIST, and Cifar-10) with varying bitwidths (i.e., 2, 4, and 6 bits). We consider two kernel sizes and two pooling schemes for each PANN model, exploiting $$2\times 2$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mn>2</mml:mn> <mml:mo>×</mml:mo> <mml:mn>2</mml:mn> </mml:mrow> </mml:math> and $$3\times 3$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mn>3</mml:mn> <mml:mo>×</mml:mo> <mml:mn>3</mml:mn> </mml:mrow> </mml:math> convolutional kernels, and max and average pooling, the latter more amenable to an optical implementation. $$3\times 3$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mn>3</mml:mn> <mml:mo>×</mml:mo> <mml:mn>3</mml:mn> </mml:mrow> </mml:math> kernels perform better than $$2\times 2$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mn>2</mml:mn> <mml:mo>×</mml:mo> <mml:mn>2</mml:mn> </mml:mrow> </mml:math> counterparts, while max and average pooling provide comparable results, with the latter performing better on MNIST and Cifar-10. The accuracy degradation due to the photonic hardware constraints is quite limited, especially on MNIST and Fashion-MNIST, demonstrating the feasibility of PANN approaches on computer vision tasks. 

/pdf/photonic-aware-neural-networks-3elnz9ps.pdf

Photonic-aware neural networks

The novel coronavirus (Covid-19) has spread rapidly amongst countries all around the globe. Compared to the rise in cases, there are few Covid-19 testing kits available. Due to the lack of testing kits for the public, it is useful to implement an automated AI-based E-health decision support system as a potential alternative method for Covid-19 detection. As per medical examinations, the symptoms of Covid-19 could be somewhat analogous to those of pneumonia, though certainly not identical. Considering the enormous number of cases of Covid-19 and pneumonia, and the complexity of the related images stored, the data pertaining to this problem of automated detection constitutes big data. With rapid advancements in medical imaging, the development of intelligent predictive and diagnostic tools have also increased at a rapid rate. Data mining and machine learning techniques are widely accepted to aid medical diagnosis. In this paper, a huge data set of X-ray images from patients with common bacterial pneumonia, confirmed Covid-19 disease, and normal healthy cases are utilized for AI-based decision support in detecting the Coronavirus disease. The transfer learning approach, which enables us to learn from a smaller set of samples in a problem and transfer the discovered knowledge to a larger data set, is employed in this study. We consider transfer learning using three different models that are pre-trained on several images from the ImageNet source. The models deployed here are VGG16, VGG19, and ResNet101. The dataset is generated by gathering different classes of images. We present our approach and preliminary evaluation results in this paper. We also discuss applications and open issues.

/pdf/transfer-learning-for-decision-support-in-covid-19-detection-43jef8nlu3.pdf

Transfer learning for decision support in Covid-19 detection from a few images in big data

Regulatory requirements for sub-sea oil and gas operators mandates the frequent inspection of pipeline assets to ensure that their degradation and damage are maintained at acceptable levels. The inspection process is usually sub-contracted to surveyors who utilize sub-sea Remotely Operated Vehicles (ROVs), launched from a surface vessel and piloted over the pipeline. ROVs capture data from various sensors/instruments which are subsequently reviewed and interpreted by human operators, creating a log of event annotations; a slow, labor-intensive and costly process. The paper presents an automatic image annotation framework that identifies/classifies key events of interest in the video footage viz. exposure, burial, field joints, anodes, and free spans. The reported methodology utilizes transfer learning with a Deep Convolutional Neural Network (ResNet-50), fine-tuned on real-life, representative data from challenging sub-sea environments with low lighting conditions, sand agitation, sea-life and vegetation. The network outputs are configured to perform multi-label image classifications for critical events. The annotation performance varies between 95.1% and 99.7% in terms of accuracy and 90.4% and 99.4% in terms of F1-Score depending on event type. The performance results are on a per-frame basis and corroborate the potential of the algorithm to be the foundation for an intelligent decision support framework that automates the annotation process. The solution can execute annotations in real-time and is significantly more cost-effective than human-only approaches.

https://www.mdpi.com/1424-8220/20/3/674/pdf

Automatic Annotation of Subsea Pipelines using Deep Learning.

Detection of tartrazine colored rice flour adulteration in turmeric from multi-spectral images on smartphone using convolutional neural network deployed on PaaS cloud

Deep learning has recently become one of the most popular sub-fields of machine learning owing to its distributed data representation with multiple levels of abstraction. A diverse range of deep learning algorithms are being employed to solve conventional artificial intelligence problems. This paper gives an overview of some of the most widely used deep learning algorithms applied in the field of computer vision. It first inspects the various approaches of deep learning algorithms, followed by a description of their applications in image classification, object identification, image extraction and semantic segmentation in the presence of noise. The paper concludes with the discussion of the future scope and challenges for construction and training of deep neural networks.

Deep Learning For Computer Vision Tasks: A review

We present the survey design, implementation, and outlook for COSMOS-Web, a 255 hour treasury program conducted by the James Webb Space Telescope in its first cycle of observations. COSMOS-Web is a contiguous 0.54 deg$^2$ NIRCam imaging survey in four filters (F115W, F150W, F277W, and F444W) that will reach 5$\sigma$ point source depths ranging $\sim$27.5-28.2 magnitudes. In parallel, we will obtain 0.19 deg$^2$ of MIRI imaging in one filter (F770W) reaching 5$\sigma$ point source depths of $\sim$25.3-26.0 magnitudes. COSMOS-Web will build on the rich heritage of multiwavelength observations and data products available in the COSMOS field. The design of COSMOS-Web is motivated by three primary science goals: (1) to discover thousands of galaxies in the Epoch of Reionization ($6<z<11$) and map reionization's spatial distribution, environments, and drivers on scales sufficiently large to mitigate cosmic variance, (2) to identify hundreds of rare quiescent galaxies at $z>4$ and place constraints on the formation of the Universe's most massive galaxies ($M_\star>10^{10}$\,M$_\odot$), and (3) directly measure the evolution of the stellar mass to halo mass relation using weak gravitational lensing out to $z\sim2.5$ and measure its variance with galaxies' star formation histories and morphologies. In addition, we anticipate COSMOS-Web's legacy value to reach far beyond these scientific goals, touching many other areas of astrophysics, such as the identification of the first direct collapse black hole candidates, ultracool sub-dwarf stars in the Galactic halo, and possibly the identification of $z>10$ pair-instability supernovae. In this paper we provide an overview of the survey's key measurements, specifications, goals, and prospects for new discovery.

Rohan Pattnaik

Papers

Deep Learning For Computer Vision Tasks: A review

COSMOS-Web: An Overview of the JWST Cosmic Origins Survey

Machine learning in astronomy