Multi-label core vector machine with a zero label

Multi-label core vector machine (Rank-CVM) is an efficient and effective algorithm for multi-label classification. But there still exist two aspects to be improved: reducing training and testing computational costs further, and detecting relevant labels effectively. In this paper, we extend Rank-CVM via adding a zero label to construct its variant with a zero label, i.e., Rank-CVMz, which is formulated as the same quadratic programming form with a unit simplex constraint and non-negative ones as Rank-CVM, and then is solved by Frank–Wolfe method efficiently. Attractively, our Rank-CVMz has fewer variables to be solved than Rank-CVM, which speeds up training procedure dramatically. Further, the relevant labels are effectively detected by the zero label. Experimental results on 12 benchmark data sets demonstrate that our method achieves a competitive performance, compared with six existing multi-label algorithms according to six indicative instance-based measures. Moreover, on the average, our Rank-CVMz runs 83 times faster and has slightly fewer support vectors than its origin Rank-CVM.     

Dynamic response of Ti3SiC2 under shock compression up to 112 GPa

Ti₃SiC₂ is of interest due to its unique dual nature reminiscent of both brittle ceramics and ductile metals at ambient conditions. In this work, plate-impact experiments have been performed to study the dynamic behavior of Ti₃SiC₂ under shock compression up to 112 GPa by using laser velocity interferometer and electric pin techniques. Hugoniot elastic limits (HEL), spall strength, and Hugoniot equations of state have been obtained based on measured particle velocity profiles and shock wave velocities. The ratio of spall strength to HEL for Ti₃SiC₂ is larger than brittle ceramics but smaller than metals. This result indicates that the dual nature of Ti₃SiC₂ remains at least up to 10 GPa. On the other hand, the linearity of the Hugoniot equation of state, $D=6.9\text{0}1(22)+1.153(53)u_{\text{p}}$, suggests that the initial structure of Ti₃SiC₂ should be stable up to 112 GPa, in contrast to the result reported by Jordan et al. [J. Appl. Phys., 93 (2003) 9639].     

Synthesis of carbon nanocapsules containing Fe, Ni or Co by arc discharge in aqueous solution

A simple, inexpensive and one-step synthesis method of metal-containing carbon nanocapsules using an arc discharge in aqueous solution is reported. It was found that Ni, Co and Fe nanoparticles could be in situ encapsulated in carbon shells when the arc was performed respectively in aqueous solutions of NiSO₄, CoSO₄ and FeSO₄. Transmission electron microscopy, energy dispersive X-ray spectroscopy and electron diffraction patterns of selected areas were used to determine the crystalline phase of the metal cores. To explain the formation mechanism of metal-containing carbon nanocapsules, a model of discharge in solution is proposed. This result presents a simply controllable way to synthesize metal-containing carbon nanocapsules.     

Precise localization of eye centers with multiple cues

Automatic detection and precise localization of human eye centers are the essential processes in photo related multimedia applications. Since eye center points are used as reference base points for further intelligent processing, precise eye center localization is very important. In face recognition the accuracy of localization of eye centers directly influences the identification accuracy. A multiple stage approach with multiple cues for detection and precise localization of eye centers is presented in this paper. Multiple scopes searching strategy is used for correctly extracting eye patch images from the background. Dedicated gradient based features and curvelet based features are constructed and used for comprehensively revealing the intensity distribution characteristics and the edge based texture around eye centers. A rebuilt score calculation mechanism is proposed and the rebuilt scores are used as a specific measurement index reflecting the matching accuracy. The final localizations of eye centers are determined with integrating the gradient based scores and curvelet based scores. The experiment results testing on public face datasets show that the localization accuracy of proposed approach outperforms the accuracy with other state of the art methods.     

Integrated size and topology optimization of skeletal structures with exact frequency constraints

The present paper studies the integrated size and topology optimization of skeletal structures under natural frequency constraints. It is found that, unlike the conventional compliance-oriented topology optimization problems, the considered problem may be strongly singular in the sense that the corresponding feasible domain may be disconnected and the global optimal solutions are often located at the tips of some separated low dimensional sub-domains when the cross-sectional areas of the structural components are used as design variables. As in the case of stress-constrained topology optimization, this unpleasant behavior may prevent the gradient-based numerical optimization algorithms from finding the true optimal topologies. To overcome the difficulties posed by the strongly singular optima, some particular forms of area/moment of inertia-density interpolation schemes, which can restore the connectedness of the feasible domain, are proposed. Based on the proposed optimization model, the probability of finding the strongly singular optimum with gradient-based algorithms can be increased. Numerical examples demonstrate the effectiveness of the proposed approach.     

广角摄像机在机器人视觉定位中的应用

为了保证广角摄像机在现代机器人视觉定位应用中的准确性,以广角摄像机像素坐标的反畸变过程为研究对象,提出一种在畸变对应表的基础上使用中心迭代法以及区域插值法确定物体实际位置的方法。通过实验证明了该方法能够测量物体的位置,其精度能够满足工业使用要求。     

Photoelectric conversion of photosynthetic reaction center in multilayered films fabricated by layer-by-layer assembly

Multilayered protein films which contained ordered layers of photosynthetic reaction center (RC) from Rhodobacter Sphaeroides (RS601) were assembled by means of alternate electrostatic adsorption with positively charged poly(diallyldimethylammonium chloride) (PDDA). The assembly of RC was monitored by spectrometry and photocurrent measurement. Linear film growth was observed up to about 20 cycles of adsorption. For the monolayer film, the photocurrent was about 8.5 nA cm⁻². For the multilayered film, the total photocurrent was about 77 nA cm⁻² for the 24-layer RC film, while the average photocurrent increment per adsorption cycle was about 3.2 nA cm⁻². The overall light-to-electricity conversion efficiency for a 24-layer film was about eight times higher than that for the monolayer one. The effects of electrode potential and pH on the photocurrent were also measured to illustrate the light-to-electric converting mechanism.