A local region based approach to lip tracking

2012 ◽  
Vol 45 (9) ◽  
pp. 3336-3347 ◽  
Author(s):  
Yiu-ming Cheung ◽  
Xin Liu ◽  
Xinge You
Keyword(s):  
1991 ◽  
Vol 3 (2) ◽  
pp. 213-225 ◽  
Author(s):  
John Platt

We have created a network that allocates a new computational unit whenever an unusual pattern is presented to the network. This network forms compact representations, yet learns easily and rapidly. The network can be used at any time in the learning process and the learning patterns do not have to be repeated. The units in this network respond to only a local region of the space of input values. The network learns by allocating new units and adjusting the parameters of existing units. If the network performs poorly on a presented pattern, then a new unit is allocated that corrects the response to the presented pattern. If the network performs well on a presented pattern, then the network parameters are updated using standard LMS gradient descent. We have obtained good results with our resource-allocating network (RAN). For predicting the Mackey-Glass chaotic time series, RAN learns much faster than do those using backpropagation networks and uses a comparable number of synapses.


2015 ◽  
Vol 282 (1816) ◽  
pp. 20152053 ◽  
Author(s):  
Terence N. Suinyuy ◽  
John S. Donaldson ◽  
Steven D. Johnson

Brood-site mutualisms represent extreme levels of reciprocal specialization between plants and insect pollinators, raising questions about whether these mutualisms are mediated by volatile signals and whether these signals and insect responses to them covary geographically in a manner expected from coevolution. Cycads are an ancient plant lineage in which almost all extant species are pollinated through brood-site mutualisms with insects. We investigated whether volatile emissions and insect olfactory responses are matched across the distribution range of the African cycad Encephalartos villosus . This cycad species is pollinated by the same beetle species across its distribution, but cone volatile emissions are dominated by alkenes in northern populations, and by monoterpenes and a pyrazine compound in southern populations. In reciprocal choice experiments, insects chose the scent of cones from the local region over that of cones from the other region. Antennae of beetles from northern populations responded mainly to alkenes, while those of beetles from southern populations responded mainly to pyrazine. In bioassay experiments, beetles were most strongly attracted to alkenes in northern populations and to the pyrazine compound in southern populations. Geographical matching of cone volatiles and pollinator olfactory preference is consistent with coevolution in this specialized mutualism.


1992 ◽  
Vol 296 ◽  
Author(s):  
Guang Gao ◽  
Ravindra Pandey ◽  
A. Barry Kunz

AbstractAn embedded molecular cluster model is used to study the electronic structure of cyclotrimethylene trinitramine(RDX) molecular crystal. In this model, a molecular cluster describing a local region of the crystal is treated in the Hartree-Fock approximation. The embedding lattice is represented by multipoles that are determined quantum-mechanically. Cluster-lattice orthogonality is achieved by transforming canonical orbitals to orbitals which are localized on individual molecular sites by means of a self-consistent localization potential. Results for the free molecule and the perfect crystal demonstrate the domination of three N-NO2 groups on intermolecular properties. For the defective crystal, results show a significant distortion in the electronic structure caused by local deformations in the lattice that may account for the formation of hot spots.


2014 ◽  
Vol 513-517 ◽  
pp. 3463-3467
Author(s):  
Li Fen Zhou ◽  
Chang Xu Cai

The Chan-Vese (C-V) active contour model has low computational complexity, initialization and insensitive to noise advantagesand utilizes global region information of images, so it is difficult to handle images with intensity inhomogeneity. The Local binary fitting (LBF) model based on local region information has its certain advantages in mages segmentation of weak boundary or uneven greay.but , the segmentation results are very sensitive to the initial contours, In order to address this problem, this paper proposes a new active contour model with a partial differential equation, which integrates both global and local region information. Experimental results show that it has a distinctive advantage over C-V model for images with intensity inhomogeneity, and it is more efficient than LBF.


2018 ◽  
Vol 10 (4) ◽  
pp. 1-13
Author(s):  
Lianghua Wen ◽  
Ping Yang ◽  
Shuai Wang ◽  
Shanqui Chen ◽  
Wenjing LIU ◽  
...  

Author(s):  
Genady P. Cherepanov

By way of introduction, the general invariant integral (GI) based on the energy conservation law is presented, with mention of cosmic, gravitational, mass, elastic, thermal and electromagnetic energy of matter application to demonstrate the approach, including Coulomb's Law generalized for moving electric charges, Newton's Law generalized for coupled gravitational/cosmic field, the new Archimedes’ Law accounting for gravitational and surface energy, and others. Then using this approach the temperature track behind a moving crack is found, and the coupling of elastic and thermal energies is set up in fracturing. For porous materials saturated with a fluid or gas, the notion of binary continuum is used to introduce the corresponding GIs. As applied to the horizontal drilling and fracturing of boreholes, the field of pressure and flow rate as well as the fluid output from both a horizontal borehole and a fracture are derived in the fluid extraction regime. The theory of fracking in shale gas reservoirs is suggested for three basic regimes of the drill mud permeation, with calculating the shape and volume of the local region of the multiply fractured rock in terms of the pressures of rock, drill mud and shale gas.


2018 ◽  
Vol 2018 ◽  
pp. 1-13 ◽  
Author(s):  
Shu Yang ◽  
Jinglin Li ◽  
Quan Yuan ◽  
Zhihan Liu ◽  
Fangchun Yang

Group sensing is a kind of crowdsensing service where HD map producers motivate private cars in a local region to collect data from real world. Group sensing needs vehicles to communicate physically and drivers to collaborate strategically in a mobile or edge-assisted environment. First, we consider collaboration module that motivates drivers to be participants; centralized and distributed motivating methods are discussed. Secondly, we consider communication module; two VANET-based methods are proposed to achieve message relaying in edge infrastructure. To accomplish participants’ selection, three combinations of two modules are proposed and simulated based on a flexible framework. The results show that centralized selection could motivate collaboration at a low price but brings heavy communication overhead. Clustered selection requires more incentives and less communication overhead than centralized selection. Distributed selection is usually the first class choice because of its fine performances on both communicating and motivating.


Sign in / Sign up

Export Citation Format

Share Document