VAViacheslav E. Antsiperov

Viacheslav E. Antsiperov

*This is an algorithmically generated author record based on publicly available information.

Viacheslav E. Antsiperov has a total of 69 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Signal Intensity Change Point Detection by System of Overlapping Receptive Fields

    by Viacheslav E. Antsiperov, Mikhail M. Gutorov, Elena R. Pavlyukova

  • Detection of abrupt intensity changes in speech signals based on the receptive field concept

    by Russia Kotelnikov Ire Ras 125009, M.M. Gutorov, V.E. Antsiperov

    This paper explores the feasibility of automatic detection of vowel and word boundaries in speech signals represented in a sampled form of a neuromorphic model of the peripheral stage of the biological auditory perception system. An algorithmic approach is proposed that performs sequential processing of speech signals to detect the temporal boundaries of speech elements. At the initial stage, a sharpness intensity index is calculated using a system of temporal receptive fields, allowing for the identification of candidate segment boundaries. This is followed by event selection for the onset and offset of speech fragments and subsequent verification of vowel boundaries based on the local average spiking activity. To enhance the detection of word boundaries, the method further incorporates envelope analysis with adaptive thresholding, ensuring robustness and reproducibility of the segmentation results. The algorithm's performance was objectively evaluated using manually annotated time-aligned vowel boundaries, enabling the calculation of root mean square error and coincidence metrics. The results demonstrate high detection accuracy, reaching precision on the order of tens of milliseconds for word boundaries, thereby confirming the practical viability of the proposed approach. The method shows elevated sensitivity at vowel offsets, occasionally resulting in false detections, which indicates the need for incorporating adaptive contextual rules. The proposed system is suitable for real-time speech signal analysis and segmentation tasks.

  • Representation of CT images using receptive fields occupation numbers with fixed-size sample counts

    by V.E. Anciperov, V.A. Kershner, G.K. Mansurov

    Modern medicine actively uses the latest visualization tools, among which the most popular are computed tomography, magnetic resonance imaging, positron emission tomography, single-photon emission computed tomography, etc. A common feature of most of them is a relatively low level of electromagnetic radiation used, which leads to the fact that the corresponding detectors form images either in the FSD mode – detectors that count photons, or in a close mode of EID – detectors that integrate the energy of photons. Due to the specificity of these modes, images formed by the corresponding medical FSD/EID visualization devices are poorly suited for traditional methods of their recognition and analysis (diagnostics). Therefore, intensive research is constantly being conducted in this area to develop new methods and approaches to processing, storing and transmitting medical images. The aim of the study is to develop methods for processing and presenting images that consider their discrete photo-counting nature. At the same time, the developed representations are focused on the possibility for a specialist to form reasonable conclusions (diagnostics) about the degree of similarity of the shape of objects in images with the shapes of previously observed objects (precedents). A new method for presenting CT images based on the sampling representation obtained by the authors has been developed. The method is based on direct modeling of the mechanisms of primary processing of video data on the periphery of the human visual system. The proposed in the work sampling representation (list-class) is very convenient in theoretical studies, but for practical applications it is more convenient to use combined representations of mixed list- and binned- classes. Experience in numerical testing and optimization of algorithms for the formation of the developed combined representation showed that it is possible to avoid computational problems associated with the processing of large amounts of data and adapt the approach to modern computer tomography problems.

  • Edge-directed image interpolation as a neuromorphic method for decoding their sampling representations

    by Viacheslav E. Antsiperov, Vladislav A. Kershner

    Subject of study. This study investigates the synthesis of data-processing and analysis procedures in the form of event streams for image encoding and decoding tasks, based on the principles of neuroiconics. Aim of study. The aims of this study are to develop a model for neuromorphic encoding and subsequent decoding of video data, synthesize procedures for image recovery, and create an algorithm for object-boundary detection in digital images. Method. A sampling representation of images is employed. A generative encoder model is constructed based on the statistics of this sampling representation, representing the joint distribution of the input and encoded data as a mixture of components. A system of receptive fields is introduced to model lateral inhibition mechanisms. Main results. The mechanisms of primary neuromorphic processing of video data in the peripheral visual system are modeled, and neuromorphic procedures for image encoding and reconstruction are synthesized. Numerical testing and optimization of the developed algorithms demonstrate the feasibility of avoiding computational difficulties associated with large-scale data processing and show the adaptability of the proposed approach to modern neural network tasks. Practical significance. The synthesized procedures can be applied to modern communication systems and related tasks such as object search, identification, and recognition in digital images. Furthermore, the proposed approach may serve as a foundation for the analysis and synthesis of other neuromorphic information systems designed to operate on data streams.

  • Decoding Neuromorphic Codes of Images in the Marr’s Paradigm

    by Viacheslav Antsiperov

    The work is devoted to the results of analysis/synthesis of image encoding based on the periphery neuromorphic model inspired by the mechanisms of human visual system. To adequately model the retinal mechanisms analysis and synthesis are carried out on the basis of the most realistic representation of input data—external stimuli—in the form of a stream of triggered receptor events. The statistical data model is chosen in the form of a Poisson random counts, the rationale for which was obtained in our previous papers. On this basis a statistical description of the input data in the form of a sampling representation is accepted. The model of retinal coding of input data is considered within the framework of the concept of receptive fields. This coding model implements a number of well-known neural mechanisms, including central/lateral inhibition. Decoding issues are considered in the context of reconstructing spatial contrasts of the image, partially modeling the responses of simple cells of the primary visual cortex. It is shown that the model of coupled ON-OFF encoding allows to restore sharp image details in the form of local edges. At the end of the article the interesting connection with the Marr’s theory of computer vision is discussed.


  • Neuromorphic Image Coding/Decoding Based on Sampling Estimates of the Local Contrast

    by V. E. Antsiperov, V. A. Kershner

    Neuromorphic methods for processing relatively large volumes of streaming data using the example of an image encoding/restoration problem are discussed. For these purposes, a special (sampling) representation of the input data is used in the form of a stream of discrete events (counts), similar to the events of discharges of neurons in the retina of the eye. Based on the specificity of sampling representation, a generative data model is naturally formalized in the form of a system of components distributed across the field of view—a system of receptive fields that embody the universal principles (including lateral inhibition) of biological neural networks. The mechanism of lateral inhibition is realized in the model in the form of an antagonistic structure of receptive fields center/surround. Issues of image decoding are considered in the context of restoration of spatial contrasts, which partly mimics the work of the so-called simple cells of the visual cortex. It is shown that the coupled ON–OFF decoding model allows for the restoration of sharp image details in the form of emphasizing contrast boundaries. The procedures of neuromorphic encoding/decoding of video data synthesized in this work can be used in modern communication systems, as well as in related tasks of searching, identifying, etc. objects in digital images.