From a theoretical point of view, Hidden Markov Models (HMMs) and Dynamic Bayesian Networks (DBNs) are similar, still in practice they pose different challenges and perform in a different manner. In this study we present a comparative analysis of the two spatial-temporal classification methods: HMMs and DBNs applied to the Facial Action Units (AUs) recognition problem. The Facial Action Coding System (FACS) developed by Ekman and Friesen decomposes the face into 46 AUs, each AU being related to the contraction of one or more specific facial muscles. FACS proved its applicability to facial behavior modeling, enabling the recognition of an extensive palette of facial expressions. Even though a lot has been published on this theme, it is still difficult to draw a conclusion regarding the best methodology to follow, as there is no common basis for comparison and sometimes no argument is given why a certain classification method was chosen. Therefore, our main contributions reside in discussing and comparing the relative performance of the two proposed classifiers (HMMs vs. DBNs) and also of different Region of Interest (ROI) selections proposed by us and different optical flow estimation methods. We can consider our automatic system towards AUs classification an important step in the facial expression recognition process, given that even one emotion can be expressed in different ways, fact that suggests the complexity of the analyzed problem. The experiments were performed on the Cohn-Kanade database and showed that under the same conditions regarding initialization, labeling, and sampling, both classification methods produced similar results, achieving the same recognition rate of 89% for the classification of facial AUs. Still, by enabling non-fixed sampling and using HTK, HMMs rendered a better performance of 93% suggesting that they are better suited for the special task of AUs recognition.
Reactive hyperemia (RH) in forearm muscle or skin microcirculation has been considered as a surrogate endpoint in clinical studies of cardiovascular disea e. We evaluated two potential confounders that might limit such use of RH, namely laterality of measurement and intake of non-steroidal anti-inflammatory drugs (NSAIDS). Twenty-three young non-smoking healthy adults were enrolled. In Experiment 1 (n=16), the RH elicited by 3 min of ischemia was recorded in the muscle (strain gauge plethysmography, hand excluded) and skin (laser Doppler imaging) of both forearms. In Experiment 2 (n=7), RH was determined in the dominant forearm only, one hour following oral acetylsalicylic acid (1 g) or placebo. In Experiment 1, peak RH was identical in both forearms, and so were the corresponding durations of responses. RH lasted significantly less in muscle than in skin (p=0.003), a hitherto unrecognized fact. In the skin, acetylsalicylate reduced duration (43 vs. 57.4 s for placebo, p=0.03), without affecting the peak response. In muscle, duration tended to decrease with acetylsalicylate (21.4 vs. 26.0 s with placebo, p=0.06) and the peak increase in blood flow was blunted (27.2 vs. 32.4 ml/min/100 ml tissue with placebo, p=0.003). We conclude that, when using RH as a surrogate endpoint in studies of cardiovascular disease, a confounding by laterality of measurement need not be feared, but NSAIDS may have an influence, although perhaps not on the peak response in the skin., G. Addor, A. Delachaux, B. Dischl, D. Hayoz, L. Liaudet, B. Waeber, F. Feihl., and Obsahuje bibliografii a bibliografické odkazy
Studied are differences of two approaches targeted to reveal latent variables in binary data. These approaches assume that the observed high dimensional data are driven by a small number of hidden binary sources combined due to Boolean superposition. The first approach is the Boolean matrix factorization (BMF) and the second one is the Boolean factor analysis (BFA). The two BMF methods are used for comparison. First is the M8 method from the BMDP statistical software package and the second one is the method suggested by Belohlavek \& Vychodil. These two are compared to BFA, especially with the Expectation-maximization Boolean Factor Analysis we had developed earlier has, however, been extended with a binarization step developed here. The well-known bars problem and the mushroom dataset are used for revealing the methods' peculiarities. In particular, the reconstruction ability of the computed factors and the information gain as the measure of dimension reduction was under scrutiny. It was shown that BFA slightly loses to BMF in performance when noise-free signals are analyzed. Conversely, BMF loses considerably to BFA when input signals are noisy.
Several counterparts of Bayesian networks based on different paradigms have been proposed in evidence theory. Nevertheless, none of them is completely satisfactory. In this paper we will present a new one, based on a recently introduced concept of conditional independence. We define a conditioning rule for variables, and the relationship between conditional independence and irrelevance is studied with the aim of constructing a Bayesian-network-like model. Then, through a simple example, we will show a problem appearing in this model caused by the use of a conditioning rule. We will also show that this problem can be avoided if undirected or compositional models are used instead.
This paper pursues the effect of changes in distance and vibration frequency on the vibration velocity amplitude. As an example, we used the vibrating sheet piles at the cons truction of a new multi-functional FEI building on the premises of VŠB - TU Ostrava, at 17 listopadu street. The effect of these changes is monitored both in in-situ measurements and in a simulated real-life situation. The calculation software Plaxis 2D is used for creation of numer ical models. At the close, the results from in-situ measurements are confronted with those achieved from the models., Tomáš Petřík, Eva Hrubešová and Markéta Lednická., and Obsahuje bibliografické odkazy
To determine whether PHEMA [poly(2-hydroxyethylmethacrylate)] is suitable for portal vein embolization in patients scheduled to right hepatectomy and whether it is as effective as the currently used agent (a histoacryl/lipiodol mixture). Two groups of nine patients each scheduled for extended right hepatectomy for primary or secondary hepatic tumor, had right portal vein embolization in an effort to induce future liver remnant (FLR) hypertrophy. One group had embolization with PHEMA, the other one with the histoacryl/lipiodol mixture. In all patients, embolization was performed using the right retrograde transhepatic access. Embolization was technically successful in all 18 patients, with no complication related to the embolization agent. Eight patients of either group developed FLR hypertrophy allowing extended right hepatectomy. Likewise, one patient in each group had recanalization of a portal vein branch. Hist ology showed that both embolization agents reach the periphery of portal vein branches, with PHEMA penetrating somewhat deeper into the periphery. PHEMA has been shown to be an agent suitable for embolization in the portal venous system comparable with existing embolization agent (histoacryl/lipiodol mixture)., J. H. Peregrin, R. Janoušek, D. Kautznerová, M. Oliverius, E. Sticová, M. Přádný, J. Michálek., and Obsahuje bibliografii
The paper deals with results of special fatigue life tests. Random processes of different power spectral densities loaded tube specimens made of a mild carbon steel ČSN 411523.1 (11523.1), notched by a perpendicular hole. It has been found that fatigue lives build similar S-N curves like by harmonic loadings, when the standard deviation s^d of peaks of an effective damaging stress and number of loading blocks Nb are used instead of amplitudes oa, and number of cycles Na, respectively. S-N curves of unaxial and multiaxial loading are compared in the paper.
The aim of this study was to investigate the effects of troglitazone (TRO) - a new insulin-sensitizing agent - on some metabolic parameters in an experimental model of hypertriglyceridemia and insulin resistance, hereditary hypertriglyceridemic rats, and to compare its effects with those of vitamin E, an antioxidant agent. Three groups of the above rats were fed diets with a high content of sucrose (70 % of energy as sucrose) for four weeks. The first group was supplemented with TRO (120 mg/kg diet), the second one with vitamin E (500 mg/kg diet), and the third group served as the control. Vitamin E supplementation did not lower serum triglycerides (2.42±0.41 vs. 3.39±0.37 mmol/l, N.S.) while TRO did (1.87±0.24 vs. 3.39±0.37 mmol/l, p<0.01). Neither TRO nor vitamin E influenced the serum levels of free fatty acids (FFA). Both drugs influenced the spectrum of fatty acids in serum phospholipids - TRO increased the levels of polyunsaturated fatty acids (PUFA) n-6 (36.04±1.61 vs. 19.65±1.56 mol %, p<0.001), vitamin E increased the levels of PUFA n-3 (13.30±0.87 vs. 6.79±0.87 mol %, p<0.001) and decreased the levels of saturated fatty acids (32.97±0.58 vs. 51.45±4.01 mol %, p<0.01). In conclusion, TRO lowered the level of serum triglycerides but vitamin E did not have this effect in hypertriglyceridemic rats. Compared with TRO, vitamin E had a different effect on the spectrum of fatty acids in serum phospholipids., Š. Chvojková, L. Kazdová, J. Divišová., and Obsahuje bibliografii
The prediction of traffic accident duration is great significant for rapid disposal of traffic accidents, especially for fast rescue of traffic accidents and re- moving traffic safety hazards. In this paper, two methods, which are based on artificial neural network (ANN) and support vector machine (SVM), are adopted for the accident duration prediction. The proposed method is demonstrated by a case study using data on approximately 235 accidents that occurred on freeways located between Dalian and Shenyang, from 2012 to 2014. The mean absolute error (MAE), the root mean square error (RMSE) and the mean absolute percentage error (MAPE) are used to evaluate the performances of the two measures. The conclusions are as follows: Both ANN and SVM models had the ability to predict traffic accident duration within acceptable limits. The ANN model gets a better result for long duration incident cases. The comprehensive performance of the SVM model is better than the ANN model for the traffic accident duration prediction.
We compare a recent selection theorem given by Chistyakov using the notion of modulus of variation, with a selection theorem of Schrader based on bounded oscillation and with a selection theorem of Di Piazza-Maniscalco based on bounded A , Λ-oscillation.