AbstractThe success of deep learning is usually accompanied by the growth in neural network depth. However, traditional training method only supervises at its last layer and propagates supervision layer-by-layer, which leads to hardship optimizing intermediate layers. Recently, has been proposed add auxiliary classifiers layers networks. By these with supervised task loss, can be applied shallo...

In this introductory chapter, Granger traces the development of Contrastive Linguistics and Translation Studies over the last decades to the present day, focusing on the role of the computer corpus in giving new impetus to each field and bringing them closer together. She discusses the different types of monolingual and multilingual corpora being used in CL and TS research, proposing at the sam...

This pape.r shows that contrastive Hebbian, the algorithm used in mean field learning, can be applied to any continuous Hopfield model. This implies that non-logistic activation functions as well as self connections are allowed. Contrary to previous approaches, the learning algorithm is derived without considering it a mean field approximation to Boltzmann machine learning. The paper includes a...

Phonological neighborhood density (ND) conditions variation in a number of acoustic phonetic properties of words. For example, previous research has shown greater hyperarticulation and greater nasal coarticulation in high-ND English words than in low-ND words. Here, we investigate the effects of ND on vowel hyperarticulation and vowel nasality (coarticulatory and contrastive) in French. Acousti...

This study aims at determining whether there are visual cues to contrastive focus in French. An audiovisual corpus was recorded from a male native speaker of French consisting of sentences with a subject-verb-object (SVO) syntactic structure. Four conditions were studied: focus on each phrase (S,V,O) and broad focus. The corpus was first acoustically validated: the pitch maximum over the uttera...

What has transpired immediately before has a strong influence on how sensory stimuli are processed and perceived. In particular, temporal context can have contrastive effects, repelling perception away from the interpretation of the context stimulus, and attractive effects (TCEs), whereby perception repeats upon successive presentations of the same stimulus. For decades, scientists have documen...

The kpml development environment ooers a comfortable graphical and mouse-click oriented user interface for developing and maintaining large-scale sets of multilingual systemic-functional linguistic descriptions. Resource sets can be tested by static integrity checks and by application in text generation mode. Particular support is given for the development of descriptions for new language on th...

CONTRASTIVE TOPIC: MEANINGS AND REALIZATIONS

This paper introduces Ranking Info Noise Contrastive Estimation (RINCE), a new member in the family of InfoNCE losses that preserves ranked ordering positive samples. In contrast to standard loss, which requires strict binary separation training pairs into similar and dissimilar samples, RINCE can exploit information about similarity ranking for learning corresponding embedding space. We show p...