Shape-from-recognition: Recognition enables meta-data transfer

Low-level cues in an image not only allow to infer higher-level information like the presence of an object, but the inverse is also true. Category-level object recognition has now reached a level of maturity and accuracy that allows to successfully feed back its output to other processes. This is what we refer to as cognitive feedback. In this paper, we study one particular form of cognitive feedback, where the ability to recognize objects of a given category is exploited to infer different kinds of meta-data annotations for images of previously unseen object instances, in particular information on 3D shape. Meta-data can be discrete, real- or vector-valued. Our approach builds on the Implicit Shape Model of Leibe and Schiele [B. Leibe, A. Leonardis, B. Schiele, Robust object detection with interleaved categorization and segmentation, International Journal of Computer Vision 77 (1–3) (2008) 259–289], and extends it to transfer annotations from training images to test images. We focus on the inference of approximative 3D shape information about objects in a single 2D image. In experiments, we illustrate how our method can infer depth maps, surface normals and part labels for previously unseen object instances.     

Designing Fast Converging Cost Sharing Methods for Multicast Transmissions

We study a multicast game in non-cooperative directed networks in which a source sends the same message or service to a set of r receiving users and the cost of the used links is divided among the receivers according to a given cost sharing method. By following the approach recently proposed by Chen et al. (Proceedings of the 19th Annual ACM-SIAM Symposium on Discrete Algorithms (SODA), pp. 854–863, 2008), we analyze the performances of a family of methods satisfying certain desiderata, namely, weak and strong budget-balance, fairness and separability. We show that any fair method may require an arbitrary number of selfish moves in order to converge to a pure Nash equilibrium, hence we focus on the solutions obtained after one round of selfish moves. We evaluate their quality according to two global social functions: the overall cost of the solution and the maximum shared cost of users. The only method satisfying all the properties is the well-known Shapley value for which we show an approximation ratio of the solutions reached after a one round walk equal to Θ(r ²). We then prove that relaxing the strong budget balance and separability properties (we call feasible any method satisfying weak budget balance and fairness) leads to improved performances since we determine a feasible method achieving an approximation ratio of the solutions reached after a one round walk equal to O(r). This bound is asymptotically optimal since we also show that any method satisfying weak budget balance cannot achieve an approximation ratio of the solutions reached after a one round walk smaller than r. Finally, we prove the NP-hardness of computing the sequence of moves leading to the best possible global performance and extend most of the results to undirected networks.     

A New Approach for Aquifer Vulnerability Assessment: the Case Study of Campania Plain

We are going to propose a new method for aquifer vulnerability assessment, named Susceptibility Index-Contamination Degree (SICODE). Starting from the assumption that soil chemistry impacts on infiltration water quality, geochemical tool such as the soil contamination degree index (CD) was combined with hydrogeological parameters in order to enhance previous well-known index (DRASTIC, Susceptibility Index). The study has been carried out at the Campania Plain (CP) aquifer, which mostly supplies the drinking water distribution system of Napoli (Southern Italy). The survey area extends from Mt. Vesuvius to the metropolitan area of Napoli and it can be considered an interesting field laboratory in order to test hydro-geochemical methods and models since both diffuse anthropogenic pollution and natural contamination sources (e.g. interaction processes between groundwater and rock) coexist. Three models have been compared. Our results have showed that DRASTIC is not the best model to be applied to urbanized environments since it does not account for the anthropogenic influence. Susceptibility Index (SI), which incorporates land use parameter, has showed a more detailed map of vulnerability degree and it better answers the local variability of human pressure. However, the proposed SICODE method completely meets the geochemical fingerprint of soil. Sensitivity analysis has revealed a high variability of the parameters due to the local heterogeneity of the analyzed system conditions. A comparison between the groundwater nitrates distribution and the predicted vulnerability has showed that. SICODE gave more accurate predictions than the other ones. This study has provided the evidence that combining hydrogeological and geochemical tools may enhance aquifer vulnerability assessment.

     

A supervised data-driven approach for microarray spot quality classification

In this paper, the problem of classifying the quality of microarray data spots is addressed, using concepts derived from the supervised learning theory. The proposed method, after extracting spots from the microarray image, computes several features, which take into account shape, color and variability. The features are classified using support vector machines, a recent statistical classification technique that is being employed widely. The proposed method does not make any assumptions on the problem and does not require any a priori information. The proposed system has been tested in a real case, for several different parameters’ configurations. Experimental results show the effectiveness of the proposed approach, also in comparison with state-of-the-art methods.     

Uptake of Escherichia coli, Vibrio cholerae non-O1 and Enterococcus durans by, and depuration of mussels (Mytilus galloprovincialis)

The uptakes of Escherichia coli, Vibrio cholerae non-O1 and Enterococcus durans by mussels (Mytilus galloprovincialis) and the times for depuration were investigated in order to determine the most useful indicator of vibrio contamination. The mussels were maintained in tanks of static seawater contaminated with bacteria at 5 log10 CFU/ml for bioaccumulation. Depuration was carried out by circulating fresh seawater through the tanks. Each organism was presented alone and with others to mussels, at temperatures of 14 and 21 degrees C. In water contaminated with either single or mixed organisms, the bacteria accumulated rapidly in the mussels reaching high concentrations after 1 h. With both single and mixed organisms, the maximum numbers of E. coli in mussels were 6.6 log10 CFU/g at 14 degrees C and 5.4 log10 CFU/g at 21 degrees C. Both V. cholerae non-O1 and E. durans alone or with other organisms reached a number ranging from 6.5 to 7 log10 CFU/g at both temperatures. During depuration the numbers of all the organisms slowly decreased, with E. coli alone, numbers ranged from 2.8 to 2 log10 CFU/g after 72 h at both 14 and 21 degrees C, and the organisms were undetectable after 144 h. With mixed organisms at 14 degrees C E. coli became undetectable after 168 h but at 21 degrees C no E. coli were recovered after 72 h. At 14 degrees C V. cholerae non-O1 alone also was undetectable after 168 h, but at 21 degrees C and with mixed organisms at both temperatures. V. cholerae was recovered after 168 h at numbers about 1 log10 CFU/g. After 168 h numbers of E. durans alone ranged from 2.6 log10 CFU/g at 14 degrees C to 1.5 log10 CFU/g at 21 degrees C, and with mixed organisms the numbers ranged from 2.3 to 2.0 log10 CFU/g at both temperatures. Of the three bacteria of faecal origin, E. durans is quickly acquired by mussels and released more slowly than the others, while E. coli quickly becomes undetectable. The results suggest that, for this kind of seafood, enterococci may be a more appropriate indicator than E. coli of risks to consumers from vibrios.     

Correlates of maximal oxygen consumption during treadmill exercise

According to the Balke treadmill protocol, 39 healthy male USAF volunteers were subjected to maximal exercise. The subjects as a group passed the anaerobic threshold by the end of exercise since average venous lactate concentrations increased from 11.2 +/- 1.6 mg% (95% confidence limits) to 93.0 +/- 8.5 mg% (95% confidence limits), and the average gas exchange ratio (R) at the end of the exercise was greater than unity (p less than 0.0005). Tests for correlations showed weak but statistically significant (p less than 0.05) relationships between change in venous lactic acid concentrations and R (r = 0.44) and maximal heart rate (r = 0.34). Maximal oxygen consumption was correlated with time of exercise (r = 0.70) and subject weight (r = 0.33). Subject age and initial plasma lactate concentrations were not significantly correlated with any other variables. Multiple linear regression yielded an equation for prediction of maximal oxygen consumption which included terms for time of exercise and subject weight. Although the multiple correlation coefficent (r = 0.75) was statistically significant (p less than 0.05), it was considered insufficient for accurate prediction of maximal oxygen consumption.     

Use of the dielectrometer for the determination of the complex dielectric constant in the microwave region in organic high polymers at various temperatures

A commercial type dielectrometer has been utilized for determining the complex dielectric constant of insulating materials as a function of temperature. For low-temperature measurements, modifications of the apparatus have been made, by which the lower section of the guide, containing the specimen, is insulated from the upper section by means of a Teflon disk, in order to permit measurements under vacuum without difficulties arising from moisture condensation. Taking into account the modifications made, relations between the experimentally measured quantities and the complex dielectric constant have been elaborated, and error analysis made, and best conditions to perform experimental measurements determined. For high- and medium-loss materials, in the low-temperature range, an accuracy of the order of 1% in ?′ and of 10% in ?″ have been calculated. Better results can be obtained for measurements at high temperature. The apparatus performance and the accuracy of measurements have been checked through measurements of the complex dielectric constant of different polar polymers at frequencies of the order of 9 × 10⁹ Hz, at temperatures between ?150 and 200°C. The experimental results are in good agreement with literature data derived from experimental measurements with other techniques and with the behavior expected on the basis of the results from radiofrequency measurements through considerations of molecular mobility in relation to molecular structure.