End-to-end neural machine translation has overtaken statistical machine translation in terms of translation quality for some language pairs, specially those with large amounts of parallel data. Besides this palpable improvement, neural networks provide several new properties. A single system can be trained to translate between many languages at almost no additional cost other than training time. Furthermore, internal representations learned by the network serve as a new semantic representation of words or sentences which, unlike standard word embeddings, are learned in an essentially bilingual or even multilingual context. In view of these properties, the contribution of the present work is two-fold. First, we systematically study the NMT context vectors, i.e. output of the encoder, and their power as an interlingua representation of a sentence. We assess their quality and effectiveness by measuring similarities across translations, as well as semantically related and semantically unrelated sentence pairs. Second, as extrinsic evaluation of the first point, we identify parallel sentences in comparable corpora, obtaining an F 1 = 98.2% on data from a shared task when using only NMT context vectors. Using context vectors jointly with similarity measures F 1 reaches 98.9%.

No abstract

This paper addresses the predictive control of the harmonic drive in an automotive application. The goal of the control was to provide good steering feel for the driver and satisfactory tracking performance in a steering system. The paper presents the dynamic model of the harmonic drive, a design framework and a two step algorithm for predictive controller design. The elaborated model predictive controller is similar to a cascade type controller with constraints in the performance function to ensure closed loop stability and a useful compromise between torque tracking and position tracking. The controller was developed and implemented in a real-time environment for rapid prototype design using Matlab, Simulink, Real-Time Workshop and dSPACE AutoBox hardware, then it was experimentally tuned for best steering feel and good tracking performance.

A cikkemben bemutatott eszköz egy olyan innovatív katalógus-vezetésre alkalmas eszköz, amelyet széles körben lehet alkalmazni. Az eszköz segítségével digitális formában lehet katalógust vezetni, ami rengeteg csalási lehetőséget zár ki. Olcsó NFC matricák/kártyák segítségével működik ez a hordozható rendszer, amelynek használata igen egyszerű. Az NFC kártyák segítségével azonosíthatják magukat a diákok. Az eszköz kiolvassa az NFC chip egyedi azonosítóját, amit egy előzőleg megtanított névvel rendel össze. Az így megkapott nevet pedig egy SD kártyára menti el egy időbélyeggel együtt.

No abstract