Pub Date : 2023-08-01DOI: 10.18287/2412-6179-co-1277
Stepan Komkov, Maksim Dzabraev, Aleksandr Petiushko
The construction of models for video action classification progresses rapidly. However, the performance of those models can still be easily improved by ensembling with the same models trained on different modalities (e.g. Optical flow). Unfortunately, it is computationally expensive to use several modalities during inference. Recent works examine the ways to integrate advantages of multi-modality into a single RGB-model. Yet, there is still room for improvement. In this paper, we explore various methods to embed the ensemble power into a single model. We show that proper initialization, as well as mutual modality learning, enhances single-modality models. As a result, we achieve state-of-the-art results in the Something-Something-v2 benchmark.
{"title":"Mutual modality learning for video action classification","authors":"Stepan Komkov, Maksim Dzabraev, Aleksandr Petiushko","doi":"10.18287/2412-6179-co-1277","DOIUrl":"https://doi.org/10.18287/2412-6179-co-1277","url":null,"abstract":"The construction of models for video action classification progresses rapidly. However, the performance of those models can still be easily improved by ensembling with the same models trained on different modalities (e.g. Optical flow). Unfortunately, it is computationally expensive to use several modalities during inference. Recent works examine the ways to integrate advantages of multi-modality into a single RGB-model. Yet, there is still room for improvement. In this paper, we explore various methods to embed the ensemble power into a single model. We show that proper initialization, as well as mutual modality learning, enhances single-modality models. As a result, we achieve state-of-the-art results in the Something-Something-v2 benchmark.","PeriodicalId":46692,"journal":{"name":"Computer Optics","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135222532","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-01DOI: 10.18287/2412-6179-co-1224
A.P. Mikitchuk, E.I. Girshova, V.V. Nikolaev
The work is devoted to an overview of the current state of optoacoustic fiber-optic ultrasonic transducers based on thermoelastic effect and fiber-optic interference receivers, its scope, technologies and materials used, the advantages and disadvantages of different methods and the prospects for the development of the industry.
{"title":"Current state of the research on optoacoustic fiber-optic ultrasonic transducers based on thermoelastic effect and fiber-optic interferometric receivers","authors":"A.P. Mikitchuk, E.I. Girshova, V.V. Nikolaev","doi":"10.18287/2412-6179-co-1224","DOIUrl":"https://doi.org/10.18287/2412-6179-co-1224","url":null,"abstract":"The work is devoted to an overview of the current state of optoacoustic fiber-optic ultrasonic transducers based on thermoelastic effect and fiber-optic interference receivers, its scope, technologies and materials used, the advantages and disadvantages of different methods and the prospects for the development of the industry.","PeriodicalId":46692,"journal":{"name":"Computer Optics","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134998065","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-01DOI: 10.18287/2412-6179-co-1276
G.E. Romanova, N.S. Nguyen
This paper presents expressions for the third-order aberrations of a Fresnel surface (Seidel coefficients). The formulas are derived in a form that allows analytical aberration analysis to be performed at the stage of layout and preliminary design of a system composed of both classical and Fresnel surfaces. In addition to the five major monochromatic Seidel aberrations of the classical surfaces and the line coma which was described for the Fresnel-type surfaces, another aberration, called quadratic astigmatism, is described in this paper. Although the obtained expressions are an approximation for the third-order aberration domain, i.e. higher-order aberrations are ignored, they provide sufficient accuracy in practice, which is also shown in the paper. The derived expressions can be applied to the analysis of aberrations in schemes using a Fresnel lens, which makes it possible to identify the areas of rational use of elements of this type.
{"title":"Third-order aberration analysis of a Fresnel lens","authors":"G.E. Romanova, N.S. Nguyen","doi":"10.18287/2412-6179-co-1276","DOIUrl":"https://doi.org/10.18287/2412-6179-co-1276","url":null,"abstract":"This paper presents expressions for the third-order aberrations of a Fresnel surface (Seidel coefficients). The formulas are derived in a form that allows analytical aberration analysis to be performed at the stage of layout and preliminary design of a system composed of both classical and Fresnel surfaces. In addition to the five major monochromatic Seidel aberrations of the classical surfaces and the line coma which was described for the Fresnel-type surfaces, another aberration, called quadratic astigmatism, is described in this paper. Although the obtained expressions are an approximation for the third-order aberration domain, i.e. higher-order aberrations are ignored, they provide sufficient accuracy in practice, which is also shown in the paper. The derived expressions can be applied to the analysis of aberrations in schemes using a Fresnel lens, which makes it possible to identify the areas of rational use of elements of this type.","PeriodicalId":46692,"journal":{"name":"Computer Optics","volume":"60 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134998066","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-01DOI: 10.18287/2412-6179-co-1239
R.E. Ilinsky
A mathematical model of the spatial distribution of the radiation flux in a multimode Hermite-Gaussian beam is proposed. In this model, the spatial distribution of the radiation flux is described by rays with radiation fluxes strung on them. A feature of the proposed model is that the radiation fluxes strung on the beams are added algebraically.
{"title":"Geometric-optical model of a multimode Hermite-Gaussian beam","authors":"R.E. Ilinsky","doi":"10.18287/2412-6179-co-1239","DOIUrl":"https://doi.org/10.18287/2412-6179-co-1239","url":null,"abstract":"A mathematical model of the spatial distribution of the radiation flux in a multimode Hermite-Gaussian beam is proposed. In this model, the spatial distribution of the radiation flux is described by rays with radiation fluxes strung on them. A feature of the proposed model is that the radiation fluxes strung on the beams are added algebraically.","PeriodicalId":46692,"journal":{"name":"Computer Optics","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134997848","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-01DOI: 10.18287/2412-6179-co-1280
Y.E. Geints, None E.K. Panina
All naturally found and technologically fabricated solid microparticles possess surface roughness. Upon optical wave scattering from such particles, in addition to its geometric shape, the surface relief becomes an important morphological factor determining the optical properties of the scatterer. We present results of the numerical 3D-simulations of focusing an optical wave with a dielectric microsphere with randomly distributed surface roughness. We address different cases of azimuthally symmetric and asymmetric distortions of the particle surface. We show that the key parameters of the near-field focal region (intensity, longitudinal and transverse dimensions) referred to as a photonic nanojet (PNJ) are sensitive to changes in the microsphere surface texture. Two important PNJ parameters, the peak intensity and the longitudinal length, are subject to more prominent changes. The influence of the optical contrast (relative refractive index) of the microsphere on PNJ parameters is investigated in detail. The possibility of reducing the influence of surface roughness on the near-field focusing strength by microsphere watering (water-uptake) is demonstrated.
{"title":"Surface roughness influence on photonic nanojet parameters of dielectric microspheres","authors":"Y.E. Geints, None E.K. Panina","doi":"10.18287/2412-6179-co-1280","DOIUrl":"https://doi.org/10.18287/2412-6179-co-1280","url":null,"abstract":"All naturally found and technologically fabricated solid microparticles possess surface roughness. Upon optical wave scattering from such particles, in addition to its geometric shape, the surface relief becomes an important morphological factor determining the optical properties of the scatterer. We present results of the numerical 3D-simulations of focusing an optical wave with a dielectric microsphere with randomly distributed surface roughness. We address different cases of azimuthally symmetric and asymmetric distortions of the particle surface. We show that the key parameters of the near-field focal region (intensity, longitudinal and transverse dimensions) referred to as a photonic nanojet (PNJ) are sensitive to changes in the microsphere surface texture. Two important PNJ parameters, the peak intensity and the longitudinal length, are subject to more prominent changes. The influence of the optical contrast (relative refractive index) of the microsphere on PNJ parameters is investigated in detail. The possibility of reducing the influence of surface roughness on the near-field focusing strength by microsphere watering (water-uptake) is demonstrated.","PeriodicalId":46692,"journal":{"name":"Computer Optics","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134998064","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-06-01DOI: 10.18287/2412-6179-co-1189
A. Pavlov, A.O. Gaugel
A task of modeling individual mental features of a decision-maker using a Fourier holography setup is considered. The problem is considered for a situation when current conditions of decision-making contradict to the previously learned rule of decision-making logic modeled by the non-cooperative game "Prisoner's Dilemma". The approach to the problem is based on a hypothesis of the correlation between mental features and the properties of the neural network as a material carrier of intelligence. The 6f Fourier holography scheme of the resonant architecture is considered as a three-layer neural network implementing a neuro-physiologically motivated concept of the "excitation ring" proposed by A.M. Ivanitsky. We analytically assess the dependence of the validity limits of the classical total probability formula for a disjunction of incompatible events on the characteristics of low-frequency filters in holograms and the correlation radii of the training image of the basic decision rule. Analytical results are confirmed by results of the numerical simulation.
{"title":"Modeling mental peculiarities of a decision maker by a Fourier-holography technique","authors":"A. Pavlov, A.O. Gaugel","doi":"10.18287/2412-6179-co-1189","DOIUrl":"https://doi.org/10.18287/2412-6179-co-1189","url":null,"abstract":"A task of modeling individual mental features of a decision-maker using a Fourier holography setup is considered. The problem is considered for a situation when current conditions of decision-making contradict to the previously learned rule of decision-making logic modeled by the non-cooperative game \"Prisoner's Dilemma\". The approach to the problem is based on a hypothesis of the correlation between mental features and the properties of the neural network as a material carrier of intelligence. The 6f Fourier holography scheme of the resonant architecture is considered as a three-layer neural network implementing a neuro-physiologically motivated concept of the \"excitation ring\" proposed by A.M. Ivanitsky. We analytically assess the dependence of the validity limits of the classical total probability formula for a disjunction of incompatible events on the characteristics of low-frequency filters in holograms and the correlation radii of the training image of the basic decision rule. Analytical results are confirmed by results of the numerical simulation.","PeriodicalId":46692,"journal":{"name":"Computer Optics","volume":"29 1","pages":""},"PeriodicalIF":1.0,"publicationDate":"2023-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83686735","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-06-01DOI: 10.18287/2412-6179-co-1226
A. Gaidel, V. Podlipnov, N. A. Ivliev, R. Paringer, P. Ishkin, S. Mashkov, R. Skidanov
Detailed automated analysis of crop images is critical to the development of smart agriculture and can significantly improve the quantity and quality of agricultural products. A hyperspectral camera potentially allows to extract more information about the observed object than a conventional one, so its use can help in solving problems that are difficult to solve with conventional methods. Often, predictive models that solve such problems require a large dataset for training. However, sufficiently large datasets of hyperspectral images of agricultural plants are not currently publicly available. Therefore, we present a new dataset of hyperspectral images of plants in this paper. This dataset can be accessed via URL https://pypi.org/project/HSI-Dataset-API/. It contains 385 hyperspectral images with a spatial resolution of 512 by 512 pixels and spectral resolution of 237 spectral bands. The images were captured in the summer of 2021 in Samara and Novocherkassk (Russia) using Offner based Imaging Hyperspectrometer of our own production. The article demonstrates the work of some basic approaches to the analysis of hyperspectral images using the dataset and states problems for further solving.
{"title":"Agricultural plant hyperspectral imaging dataset","authors":"A. Gaidel, V. Podlipnov, N. A. Ivliev, R. Paringer, P. Ishkin, S. Mashkov, R. Skidanov","doi":"10.18287/2412-6179-co-1226","DOIUrl":"https://doi.org/10.18287/2412-6179-co-1226","url":null,"abstract":"Detailed automated analysis of crop images is critical to the development of smart agriculture and can significantly improve the quantity and quality of agricultural products. A hyperspectral camera potentially allows to extract more information about the observed object than a conventional one, so its use can help in solving problems that are difficult to solve with conventional methods. Often, predictive models that solve such problems require a large dataset for training. However, sufficiently large datasets of hyperspectral images of agricultural plants are not currently publicly available. Therefore, we present a new dataset of hyperspectral images of plants in this paper. This dataset can be accessed via URL https://pypi.org/project/HSI-Dataset-API/. It contains 385 hyperspectral images with a spatial resolution of 512 by 512 pixels and spectral resolution of 237 spectral bands. The images were captured in the summer of 2021 in Samara and Novocherkassk (Russia) using Offner based Imaging Hyperspectrometer of our own production. The article demonstrates the work of some basic approaches to the analysis of hyperspectral images using the dataset and states problems for further solving.","PeriodicalId":46692,"journal":{"name":"Computer Optics","volume":"80 1","pages":""},"PeriodicalIF":1.0,"publicationDate":"2023-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76100992","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-06-01DOI: 10.18287/2412-6179-co-1247
S. Stafeev, V. D. Zaitcev, V. Kotlyar
In this paper, using the Richards-Wolf equations, we analyze focusing circularly polarized light with flat diffractive lenses. It is shown that as the numerical aperture of the lens increases, the size of the focal spot first decreases and then begins to grow. The minimum focal spot is observed at NA=0.96 (FWHM=0.55λ). With a further increase in the numerical aperture of the lens, the growth of the longitudinal component leads to an increase in the size of the focal spot. When the flat diffractive lens is replaced by an aplanatic lens, the size of the focal spot decreases monotonically as the numerical aperture of the lens increases.
{"title":"Minimal focal spot obtained by focusing circularly polarized light","authors":"S. Stafeev, V. D. Zaitcev, V. Kotlyar","doi":"10.18287/2412-6179-co-1247","DOIUrl":"https://doi.org/10.18287/2412-6179-co-1247","url":null,"abstract":"In this paper, using the Richards-Wolf equations, we analyze focusing circularly polarized light with flat diffractive lenses. It is shown that as the numerical aperture of the lens increases, the size of the focal spot first decreases and then begins to grow. The minimum focal spot is observed at NA=0.96 (FWHM=0.55λ). With a further increase in the numerical aperture of the lens, the growth of the longitudinal component leads to an increase in the size of the focal spot. When the flat diffractive lens is replaced by an aplanatic lens, the size of the focal spot decreases monotonically as the numerical aperture of the lens increases.","PeriodicalId":46692,"journal":{"name":"Computer Optics","volume":"227 1","pages":""},"PeriodicalIF":1.0,"publicationDate":"2023-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80177277","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-06-01DOI: 10.18287/2412-6179-co-1216
I. Bychkov, G. M. Ruzhnikov, R. Fedorov, A. K. Popova, Y. V. Avramenko
Various combinations of neural network parameters and sets of input data for satellite image classification are considered in the article. The training set is completed with a NDVI (normalized difference vegetation index) and local binary patterns. Testing of classifiers created on a different number of epochs and samples is carried out. Values of the neural network hyperparameters are determined that allow a classification accuracy of 0.70 and an F-measure of 0.65 to be achieved. Separation into classes with similar spectral characteristics is shown to offer low classification quality at different parameters and input data sets. Additional information is required. For example, for forests to be divided into more detailed classes, one needs to employ classifiers that use images from different seasons and vegetation periods. In addition, the training set needs to be extended to take into account various natural zones, soils, etc.
{"title":"On classification of Sentinel-2 satellite images by a neural network ResNet-50","authors":"I. Bychkov, G. M. Ruzhnikov, R. Fedorov, A. K. Popova, Y. V. Avramenko","doi":"10.18287/2412-6179-co-1216","DOIUrl":"https://doi.org/10.18287/2412-6179-co-1216","url":null,"abstract":"Various combinations of neural network parameters and sets of input data for satellite image classification are considered in the article. The training set is completed with a NDVI (normalized difference vegetation index) and local binary patterns. Testing of classifiers created on a different number of epochs and samples is carried out. Values of the neural network hyperparameters are determined that allow a classification accuracy of 0.70 and an F-measure of 0.65 to be achieved. Separation into classes with similar spectral characteristics is shown to offer low classification quality at different parameters and input data sets. Additional information is required. For example, for forests to be divided into more detailed classes, one needs to employ classifiers that use images from different seasons and vegetation periods. In addition, the training set needs to be extended to take into account various natural zones, soils, etc.","PeriodicalId":46692,"journal":{"name":"Computer Optics","volume":"36 1","pages":""},"PeriodicalIF":1.0,"publicationDate":"2023-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86902367","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
M. A. Pavlova, V. Timofeev, D. Bocharov, D. Sidorchuk, A. L. Nurmukhametov, A. Nikonorov, M.S. Yarykina, I. Kunina, A. Smagina, M. Zagarev
This paper considers an issue of delineating agricultural fields in satellite images. In this task we follow a multi-temporal data approach. We show that on such data, good quality can be achieved using a simple low-parameter method. The method consists of a combination of a field detector and an edge detector. The field detection is based on an Otsu thresholding technique and for the edge detection we use a Canny detector. Facing a lack of available datasets and aiming to estimate the proposed algorithm, we prepared and published our dataset consisting of 18,859 expertly annotated fields using Sentinel-2 data. We implement one of the state-of-the-art deep-learning approaches and compare it with the proposed method on our dataset. The experiment shows the proposed simple multi-temporal algorithm to outperform the state-of-the-art instant data approach. This result confirms the importance of using multi-temporal data and for the first time demonstrates that the delineation problem can be solved at a lower cost without loss of quality. Our approach requires a significantly less amount of training data when compared with the NN-based one. The dataset of agricultural fields used in the work and the proposed algorithm implementation in Python are published in open access.
{"title":"Low-parameter method for delineation of agricultural fields in satellite images based on multi-temporal MSAVI2 data","authors":"M. A. Pavlova, V. Timofeev, D. Bocharov, D. Sidorchuk, A. L. Nurmukhametov, A. Nikonorov, M.S. Yarykina, I. Kunina, A. Smagina, M. Zagarev","doi":"10.18287/-6179-co-1235","DOIUrl":"https://doi.org/10.18287/-6179-co-1235","url":null,"abstract":"This paper considers an issue of delineating agricultural fields in satellite images. In this task we follow a multi-temporal data approach. We show that on such data, good quality can be achieved using a simple low-parameter method. The method consists of a combination of a field detector and an edge detector. The field detection is based on an Otsu thresholding technique and for the edge detection we use a Canny detector. Facing a lack of available datasets and aiming to estimate the proposed algorithm, we prepared and published our dataset consisting of 18,859 expertly annotated fields using Sentinel-2 data. We implement one of the state-of-the-art deep-learning approaches and compare it with the proposed method on our dataset. The experiment shows the proposed simple multi-temporal algorithm to outperform the state-of-the-art instant data approach. This result confirms the importance of using multi-temporal data and for the first time demonstrates that the delineation problem can be solved at a lower cost without loss of quality. Our approach requires a significantly less amount of training data when compared with the NN-based one. The dataset of agricultural fields used in the work and the proposed algorithm implementation in Python are published in open access.","PeriodicalId":46692,"journal":{"name":"Computer Optics","volume":"85 1","pages":""},"PeriodicalIF":1.0,"publicationDate":"2023-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88231992","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：