We introduce a novel dataset for visual recognition systems in retail automation, focusing specifically on fruits and vegetables. The dataset comprises 34 species and 65 varieties, featuring fairly balanced classes and including packed goods in plastic bags. We capture each sample from multiple viewpoints and provide additional annotations, such as object count and total weight. Furthermore, a subset of samples for each class includes segmentation masks. This dataset aims to overcome the limitations of current open-access datasets by providing a more comprehensive and diverse set of training data. A total of 72 annotators collected over 100,000 images of 370,000 objects across multiple shops and cities. Around 9,000 images have manual segmentation masks. To facilitate research in this area, we provide baseline results for zero-shot and supervised classification, instance segmentation, and object counting tasks. We also investigate the impact of packaging and background type on model performance. Ultimately, this dataset is designed to support the development of multitask models for visual recognition in offline retail settings.
A comparative analysis has been made of the superluminescence spectra of two HPHT diamond samples with NV¯-centers with the concentration in the growth sectors 111 of NV¯-centers ≈10 ppm and of C-centers ≈ 45 ppm. A broad structureless superluminescence band with a maximum about 706 nm and a narrow band in the region of 705 to 720 nm were identified in the spectral range of 650–750 nm when excited by pulsed laser radiation at λ = 532 nm. It was found that the narrow-band superluminescence was not excited by pumping with continuous radiation at λ = 532 nm and the intensities of both types of superluminescence were summated under simultaneous pumping by continuous and pulsed radiation at λ = 532 nm. These data indicate either the different nature of the two types of superluminescence or that they belonged to NV¯ centers but with different excitation mechanisms. The existence of connection between the shape of the spectra of narrow-band superluminescence and of radiation pulses has been demonstrated. The observed features of narrow-band superluminescence are explained for the first time by the occurrence of mode-locking laser generation. The mode-locking laser generation without the use of reflective coatings on the sample or an external resonator on other laser materials has not been observed before.
In this paper, we solve the problem of assigning the desired characteristic polynomial of a linear time-invariant single-input dynamical system with output dynamic feedback in the form of a first-order dynamic compensator. Necessary and sufficient conditions for the existence of the solution to the problem are considered. An explicit formula for the compensator parameters, analogous to the Bass–Gura formula for a state feedback system, is derived.
In our work, we consider the Manakov system and the monodromy matrix associated with it, which plays a key role in the construction of multiphase solutions. All integrable nonlinear equations from the hierarchy of the Manakov system are expressed in terms of elements of this monodromy matrix. The spectral curves of the multiphase solutions of each of the equations from this hierarchy are determined by the characteristic equation of the monodromy matrix. The stationary equations, which are satisfied by multiphase solutions of the evolutionary hierarchy equations, can also be written using elements of the monodromy matrix. In the present work, the simplest nontrivial stationary equations are considered and solutions are constructed. These solutions are expressed in terms of integrals from solutions to Fuchsian equations. Depending on the values of parameters of the stationary equations, these Fuchsian equations can have five, four, or three singular points. The values of the parameters at which the solutions are expressed through elliptic or elementary functions are found. In the case of Fuchsian equations with three singular points (hypergeometric equations), the solutions to the Manakov system have the form of positons. The coefficients of equations of the corresponding spectral curves are found for all the considered solutions. In the cases where the solutions to the Manakov system can be classified as positons, nonhyperelliptic spectral curves have both twofold branching points and simple points. This fact has a direct correspondence with the properties of the spectral curves of positons of those integrable nonlinear equations whose spectral curves are hyperelliptic.
The paper is devoted to the problem of pattern decomposition in solving the problem of pattern recognition. The problem of pattern decomposition is considered regardless of the recognition algorithms used. The only requirement consists in the fact that the pattern recognition problem should have a classical formulation. It is shown that, with no reference to the knowledge model, pattern decomposition cannot be performed in terms of the scope of the recognition problem itself, as this leads to a revision of this recognition problem. In the same cases, when the pattern recognition problem is retained in the course of decomposition, it can change in such a way that its solution in the decomposed form is not identical to the solution to the initial pattern recognition problem.
