Optimal Sequential Grouping for Robust Video Scene Detection Using Multiple Modalities

2017 ◽  
Vol 11 (02) ◽  
pp. 193-208 ◽  
Author(s):  
Daniel Rotman ◽  
Dror Porat ◽  
Gal Ashour

Video scene detection is the task of dividing a video into semantic sections. To perform this fundamental task, we propose a novel and effective method for temporal grouping of scenes using an arbitrary set of features computed from the video. We formulate the task of video scene detection as a generic optimization problem to optimally group shots into scenes, and propose an efficient procedure for solving the optimization problem based on a novel dynamic programming scheme. This unique formulation directly results in a temporally consistent segmentation, and has the advantage of being parameter-free, making it applicable across various domains. We provide detailed experimental results, showing that our algorithm outperforms current state-of-the-art methods. To assess the comprehensiveness of this method even further, we present experimental results testing different types of modalities and their applicability in this formulation.

Electronics ◽  
2021 ◽  
Vol 10 (6) ◽  
pp. 741
Author(s):  
Yuseok Ban ◽  
Kyungjae Lee

Many researchers have suggested improving the retention of a user in the digital platform using a recommender system. Recent studies show that there are many potential ways to assist users to find interesting items, other than high-precision rating predictions. In this paper, we study how the diverse types of information suggested to a user can influence their behavior. The types have been divided into visual information, evaluative information, categorial information, and narrational information. Based on our experimental results, we analyze how different types of supplementary information affect the performance of a recommender in terms of encouraging users to click more items or spend more time in the digital platform.


2021 ◽  
Vol 13 (2) ◽  
pp. 973
Author(s):  
Gigel Paraschiv ◽  
Georgiana Moiceanu ◽  
Gheorghe Voicu ◽  
Mihai Chitoiu ◽  
Petru Cardei ◽  
...  

Our paper presents the hammer mill working process optimization problem destined for milling energetic biomass (MiscanthusGiganteus and Salix Viminalis). For the study, functional and constructive parameters of the hammer mill were taken into consideration in order to reduce the specific energy consumption. The energy consumption dependency on the mill rotor spinning frequency and on the sieve orifices in use, as well as on the material feeding flow, in correlation with the vegetal biomass milling degree was the focus of the analysis. For obtaining this the hammer mill was successively equipped with 4 different types of hammers that grind the energetic biomass, which had a certain humidity content and an initial degree of reduction ratio of the material. In order to start the optimization process of hammer mill working process, 12 parameters were defined. The objective functions which minimize hammer mill energy consumption and maximize the milled material percentage with a certain specific granulation were established. The results obtained can serve as the basis for choosing the optimal working, constructive, and functional parameters of hammer mills in this field, and for a better design of future hammer mills.


2011 ◽  
Vol 11 (02) ◽  
pp. 215-236 ◽  
Author(s):  
MATTEO BROGGI ◽  
ADRIANO CALVI ◽  
GERHART I. SCHUËLLER

Cylindrical shells under axial compression are susceptible to buckling and hence require the development of enhanced underlying mathematical models in order to accurately predict the buckling load. Imperfections of the geometry of the cylinders may cause a drastic decrease of the buckling load and give rise to the need of advanced techniques in order to consider these imperfections in a buckling analysis. A deterministic buckling analysis is based on the use of the so-called knockdown factors, which specifies the reduction of the buckling load of the perfect shell in order to account for the inherent uncertainties in the geometry. In this paper, it is shown that these knockdown factors are overly conservative and that the fields of probability and statistics provide a mathematical vehicle for realistically modeling the imperfections. Furthermore, the influence of different types of imperfection on the buckling load are examined and validated with experimental results.


Author(s):  
Alexandr Klimchik ◽  
Anatol Pashkevich ◽  
Stéphane Caro ◽  
Damien Chablat

The paper focuses on the extension of the virtual-joint-based stiffness modeling technique for the case of different types of loadings applied both to the robot end-effector and to manipulator intermediate points (auxiliary loading). It is assumed that the manipulator can be presented as a set of compliant links separated by passive or active joints. It proposes a computationally efficient procedure that is able to obtain a non-linear force-deflection relation taking into account the internal and external loadings. It also produces the Cartesian stiffness matrix. This allows to extend the classical stiffness mapping equation for the case of manipulators with auxiliary loading. The results are illustrated by numerical examples.


1979 ◽  
Vol 57 (4) ◽  
pp. 400-403 ◽  
Author(s):  
Anne Le Narvor ◽  
Pierre Saumagne

The ir spectra of mixtures of methyl propionate/water and methyl propionate/Ba2+ in dimethylsulfoxide and in acetonitrile have been recorded in the region of the νCO mode of the ester. Evidence is presented to indicate the presence of different types of complexes; their concentration was determined as a function of the composition of the medium. The spectroscopic results are compared to those from the kinetics of the alkaline hydrolysis in the same conditions. It is demonstrated that the orbital control explains the experimental results better than does the charge density on the carbon of the carbonyl group. [Journal translation]


1950 ◽  
Vol 17 (2) ◽  
pp. 145-153 ◽  
Author(s):  
J. O. Hinze ◽  
H. Milborn

Abstract Liquid, supplied through a stationary tube to the inner part of a rotating cup widening toward a brim, flows viscously in a thin layer toward this brim and is then flung off, all by centrifugal action. The flow within this layer and the disintegration phenomena occurring beyond the brim have been studied, experimentally as well as theoretically. A formula has been derived for the thickness and for the radial velocity of the liquid layer within the cup, which proved to agree reasonably well with experimental results. Three essentially different types of disintegration may take place around and beyond the edge of the cup designated, respectively, by: (a) the state of direct drop formation; (b) the state of ligament formation; and (c) the state of film formation. Which one of these is realized depends upon working conditions. Transition from state (a) into (b), or of state (b) into state (c) is promoted by an increased quantity of supply, an increased angular speed, a decreased diameter of the cup, an increased density, an increased viscosity, and a decreased surface tension of the liquid. The experimental results have been expressed in relationships between relevant dimensionless groups. For the state of ligament formation a semiempirical relationship has been derived between the number of ligaments and dimensionless groups determining the working conditions of the cup. Results of drop-size measurements made for the state of ligament formation as well as for the state of film formation show that atomization by mere rotation of the cup is much more uniform than commonly achieved with pressure atomizers.


Author(s):  
Rapeepan Promyoo ◽  
Hazim El-Mounayri ◽  
Kody Varahramyan

Atomic force microscopy (AFM) has been widely used for nanomachining and fabrication of micro/nanodevices. This paper describes the development and validation of computational models for AFM-based nanomachining. Molecular Dynamics (MD) technique is used to model and simulate mechanical indentation at the nanoscale for different types of materials, including gold, copper, aluminum, and silicon. The simulation allows for the prediction of indentation forces at the interface between an indenter and a substrate. The effects of tip materials on machined surface are investigated. The material deformation and indentation geometry are extracted based on the final locations of the atoms, which have been displaced by the rigid tool. In addition to the modeling, an AFM was used to conduct actual indentation at the nanoscale, and provide measurements to which the MD simulation predictions can be compared. The MD simulation results show that surface and subsurface deformation found in the case of gold, copper and aluminum have the same pattern. However, aluminum has more surface deformation than other materials. Two different types of indenter tips including diamond and silicon tips were used in the model. More surface and subsurface deformation can be observed for the case of nanoindentation with diamond tip. The indentation forces at various depths of indentation were obtained. It can be concluded that indentation force increases as depth of indentation increases. Due to limitations on computational time, the quantitative values of the indentation force obtained from MD simulation are not comparable to the experimental results. However, the increasing trends of indentation force are the same for both simulation and experimental results.


Sign in / Sign up

Export Citation Format

Share Document