Search

Extreme Learning Machines (ELM) have been successfully applied to variety of classification problems by utilizing a single descriptor type. However, a single descriptor may be insufficient for the visual classification task, due to the high level of intra-class variability coupled with low inter-class distance. Although several studies have investigated methods for combining multiple descriptors by ELM, they predominantly apply a simple concatenation of descriptors before classifying them. This type of descriptor fusion may impose problems of descriptor compatibility, high dimensionality and restricted accuracy. In this paper, we propose a hierarchical descriptors fusion strategy at the decision level (“late-fusion”), which relies on ELM ensembles (ELM-E). The proposed method, denoted as H-ELM-E, effectively combines multiple complementary descriptors by a two-level ELM-E based architecture, which ensures that a more informative descriptors will gain more impact on the final decision. In the first level, a separate ELM-E classifier is trained for every image descriptor. In the second level, the output scores from the previous level are aggregated into the mid-level representation which is conducted to an additional ELM-E classifier. The exhaustive experimental evaluation confirmed that the proposed hierarchical ELM-E based strategy is superior to the single-descriptor methods as well as “early fusion” of multiple descriptors, for the visual classification task. Additionally, it was shown that significant accuracy improvement is achieved by integrating ensembles of ELM as a basic classifier, instead of using a single ELM.

A new kind of the allpole lowpass filter functions with Chebyshev behaviour in the passband is derived by minimizing the passband insertion loss. The ripple parameter e , commonly used to control the attenuation at the cutoff frequency, in proposed approximation is used as a degree of freedom by which the passband attenuation can be controlled, while keeping the attenuation at the cutoff frequency equal to 3 dB. The approximation method relies on solving the nonlinear equation such that the area under the filter characteristic function in passband is minimized. Obtained filters, called Optimum Chebyshev filters, exhibit minimum insertion loss and maximum cutoff slope. Filter polynomial coefficients and the elements values for the LC ladder implementation are tabulated for filter degrees from 3 to 11. Comparison of the Optimum Chebyshev and Least Square Monotonic filters is also discussed in the paper, since both filters are developed subject to the minimum reflected power in the passband; the first with the equiripple and the second with the staircase (monotonic) behavior in the passband. It is also revealed that the Optimum Chebyshev approximation can be considered as the optimal solution among the class of filters whose passband response is bound to be a staircase.

We present a method for real-time scene classification which achieves high accuracy without a time consuming descriptor learning step and kernelized classifiers. Robustness of the classification is achieved by combining the powerful multi-channel Gabor-based descriptors and an ensemble of Extreme Learning Machines (ELM). We first extend the recently introduced Binary Gabor Patterns (BGP) to multi-channel images. This is done by extracting BGP over several color channels and embedding an additional compact color layout descriptor. Then we propose an effective method for the aggregation of multiple ELMs into a single classification system, which leads to significant classification accuracy improvements. The experimental evaluation demonstrates that multi-channel color information constantly improves classification results. The integration of multiple ELMs into an ensemble using the proposed aggregation strategy significantly outperforms linear SVM in terms of accuracy, and reaches results similar to the non-linear SVM while operating in real time. Therefore, an ensemble of ELMs with the proposed aggregation strategy could be used as an efficient alternative to the non-linear SVM for remote sensing image classification tasks.

A new procedure for design of linear phase IIR notch filters is presented in this paper. Resulted filter is realized using parallel connection of two IIR all-pass filters with approximately linear phase. Adequate amplitude characteristic is obtained by approximation of all-pass filter phase. Amplitude characteristic satisfies all prescribed demands but it is not equiripple no matter both all-pass filters have equiripple phase. Algorithm exhibits fast convergence and easy initial values determination.

Gastrectomy often results in increased likelihood of osteoporosis, metabolic aberration, and risk of fracture, and there is a need for a gastric peptide with osteogenic activity. A novel stomach pentadecapeptide, BPC-157, improves wound and fracture healing in rats in addition to having an angiogenic effect. Therefore, in the present study, using a segmental osteoperiosteal bone defect (0.8 cm, in the middle of the left radius) that remained incompletely healed in all control rabbits for 6 weeks (assessed in 2 week intervals), pentadecapeptide BPC-157 was further studied (either percutaneously given locally [10 microg/kg body weight] into the bone defect, or applied intramuscularly [intermittently, at postoperative days 7, 9, 14, and 16 at 10 microg/kg body weight] or continuously [once per day, postoperative days 7-21 at 10 microg or 10 ng/kg body weight]). For comparison, rabbits percutaneously received locally autologous bone marrow (2 mL, postoperative day 7). As standard treatment, immediately after its formation, the bone defect was filled with an autologous cortical graft. Saline-treated (2 mL intramuscularly [i.m.] and 2 mL locally into the bone defect), injured animals were used as controls. Pentadecapeptide BPC-157 significantly improved the healing of segmental bone defects. For instance, upon radiographic assessment, the callus surface, microphotodensitometry, quantitative histomorphometry (10 microg/kg body weight i.m. for 14 days), or quantitative histomorphometry (10 ng/kg body weight i.m. for 14 days) the effect of pentadecapeptide BPC-157 was shown to correspond to improvement after local application of bone marrow or autologous cortical graft. Moreover, a comparison of the number of animals with unhealed defects (all controls) or healed defects (complete bony continuity across the defect site) showed that besides pentadecapeptide intramuscular application for 14 days (i.e., local application of bone marrow or autologous cortical graft), also following other pentadecapeptide BPC-157 regimens (local application, or intermittent intramuscular administration), the number of animals with healed defect was increased. Hopefully, in the light of the suggested stomach significance for bone homeostasis, the possible relevance of this pentadecapeptide BPC-157 effect (local or intramuscular effectiveness, lack of unwanted effects) could be a basis for methods of choice in the future management of healing impairment in humans, and requires further investigation.