GWCNN: A Metric Alignment Layer for Deep Shape Analysis

Deep neural networks provide a promising tool for incorporating semantic information in geometry processing applications. Unlike image and video processing, however, geometry processing requires handling unstructured geometric data, and thus data representation becomes an important challenge in this framework. Existing approaches tackle this challenge by converting point clouds, meshes, or polygon soups into regular representations using, e.g., multi‐view images, volumetric grids or planar parameterizations. In each of these cases, geometric data representation is treated as a fixed pre‐process that is largely disconnected from the machine learning tool. In contrast, we propose to optimize for the geometric representation during the network learning process using a novel metric alignment layer. Our approach maps unstructured geometric data to a regular domain by minimizing the metric distortion of the map using the regularized Gromov–Wasserstein objective. This objective is parameterized by the metric of the target domain and is differentiable; thus, it can be easily incorporated into a deep network framework. Furthermore, the objective aims to align the metrics of the input and output domains, promoting consistent output for similar shapes. We show the effectiveness of our layer within a deep network trained for shape classification, demonstrating state‐of‐the‐art performance for nonrigid shapes.     

Blends of thermoplastic polyurethane and polypropylene. I. Mechanical and phase behavior

Pure thermoplastic polyurethane (TPU), polypropylene (PP), and TPU/PP blends with different weight ratios prepared in a twin‐screw extruder were investigated by dynamic mechanical analysis (DMA), the universal tester for mechanical investigation, and by wide‐angle X‐ray diffraction (WAXD). The addition of PP above 20 wt % to the TPU stepwise changed the ductility and Young's modulus, i.e., apparently a kind of ductile → brittle transition occurred between TPU/PP 80/20 and TPU/PP 60/40 blends. This fact and the result of analysis of WAXD curves indicated matrix → dispersed phase inversion in this concentration region. TPU melt enabled easier migration of the PP chains and prolonged crystallization of PP matrix during solidification process affecting thus crystallite size, orientation, and crystallinity. In accordance to this fact, DMA results indicated partial miscibility of PP with polyurethane in the TPU/PP blends due to the lack of interfacial interaction and adhesion between the nonpolar crystalline PP and polar TPU phases. © 2007 Wiley Periodicals, Inc. J Appl Polym Sci 104, 3980–3985, 2007     

Collaborative Writing of Text and Hypertext

A good writer understands his audience. Collaborative writing allows authors to act as readers and thus to create text which an audience is more likely to appreciate. This paper describes four sets of experiences in collaborative writing. In one of the experiences secondary school students who were unable to produce a sophisticated document independently were able to produce a quality document through a group process. Given the correct combination of authors and goals, a simple computer editor can be very helpful. Hypertext differs from text in that links among chunks of text are made explicit. It could be argued that collaborative writing would work particularly well with hypertext because the links among the writers' ideas could be more easily explored. In experiences with graduate students at two universities, this hypothesis has not been supported. The students did not understand how these links should be created and thus could not collaboratively write hypertext.     

An Optimization Approach to Improving Collections of Shape Maps

Finding an informative, structure‐preserving map between two shapes has been a long‐standing problem in geometry processing, involving a variety of solution approaches and applications. However, in many cases, we are given not only two related shapes, but a collection of them, and considering each pairwise map independently does not take full advantage of all existing information. For example, a notorious problem with computing shape maps is the ambiguity introduced by the symmetry problem — for two similar shapes which have reflectional symmetry there exist two maps which are equally favorable, and no intrinsic mapping algorithm can distinguish between them based on these two shapes alone. Another prominent issue with shape mapping algorithms is their relative sensitivity to how “similar” two shapes are — good maps are much easier to obtain when shapes are very similar. Given the context of additional shape maps connecting our collection, we propose to add the constraint of global map consistency, requiring that any composition of maps between two shapes should be independent of the path chosen in the network. This requirement can help us choose among the equally good symmetric alternatives, or help us replace a “bad” pairwise map with the composition of a few “good” maps between shapes that in some sense interpolate the original ones. We show how, given a collection of pairwise shape maps, to define an optimization problem whose output is a set of alternative maps, compositions of those given, which are consistent, and individually at times much better than the original. Our method is general, and can work on any collection of shapes, as long as a seed set of good pairwise maps is provided. We demonstrate the effectiveness of our method for improving maps generated by state‐of‐the‐art mapping methods on various shape databases.     

Milk-based beverages obtained by Kombucha application

This paper investigates the manufacture of milk-based beverages by application of several Kombucha starters. Local Kombucha culture was grown up on three substrates: sweetened black and green tea, and topinambur. Their concentrates were obtained by vacuum-evaporation and amounts of 10% and 15% (v/v) were applied to milk (2.2% fat). The traditional yoghurt starter (B3) was applied for producing control samples. All fermentations were stopped when the pH reached 4.4. Fermentation curves were registered, linear for yoghurt and sigmoidal for Kombucha. Two times faster process was achieved with yoghurt starter. Influence of inoculum concentration on the rate of fermentation was insignificant. Viscosities were higher for Kombucha beverages at lower speeds of spindle, but lower at higher speeds of spindle. Very high sensory scores were achieved for all beverages, after production and after 5-days’ storage.     

Effect of passive ultrassonic instrumentation as a final irrigation protocol on debris and smear layer removal—a sem analysis

This study sought to evaluate the efficacy of passive ultrasonic irrigation (PUI) on removing the smear layer and debris from root dentin using scanning electron microscopy (SEM). Twenty‐five bovine incisors were manually prepared and divided into three groups according to the final irrigation protocol: EDTA, final irrigation with 12 mL of 17% EDTA for 3 minutes followed by 5 mL of 2.5% NaOCl; EDTA/PUI, final flush with 4 mL of 17% EDTA and PUI for 30 seconds. These procedures were repeated three times to standardize the volume of the irrigant. Control group, after preparation, the specimens were irrigated only with 17 mL of 2.5% NaOCl. The roots were fractured and analyzed using SEM. The intragroup analysis revealed that the EDTA/PUI protocol removed a higher amount of debris at the cervical third (P = 0.03). The intergroup analysis revealed that EDTA/PUI presented the lowest amount of debris at the cervical third (P = 0.007). Smear layer scores were higher in the control group compared with the EDTA and EDTA/PUI groups, but only at the cervical third (P = 0.02). None of the final irrigant protocols completely removed the smear layer and debris. EDTA/PUI only improved the removal of debris at the cervical third. Microsc. Res. Tech. 76:496–502, 2013. © 2013 Wiley Periodicals, Inc.