This paper aims to improve the inaccuracy problem of the existing informatized caption in the noisy environment by using the additional caption information. The IBM Watson API can automatically generate the informatized caption including the timing information and the speaker ID information from the voice information input. In this IBM Watson API, when there is noise in the voice signal, the re...

Visual Question Answering (VQA) is the task of taking as input an image and a free-form natural language question about the image, and producing an accurate answer. In this work we view VQA as a “feature extraction” module to extract image and caption representations. We employ these representations for the task of image-caption ranking. Each feature dimension captures (imagines) whether a fact...

Section header conforming to settings and italics body. figure no longer floating because that would not make sense in a poster. table no longer floating. algorithm no longer floating (note: do not include algorithm.sty). As a consequence of the redefinition of figure the command \subfigure has been redefined. Therefore, do not include package subfigure.sty. All redefinitions adhere to the orig...

We present a nonparametric density estimation technique for image caption generation. Data-driven matching methods have shown to be effective for a variety of complex problems in Computer Vision. These methods reduce an inference problem for an unknown image to finding an existing labeled image which is semantically similar. However, related approaches for image caption generation (Ordonez et a...

This study investigates the effects of different display modes of video captions on mobile devices, including non-caption, full-caption, and target-word modes, on the English comprehension and vocabulary acquisition of fifth graders. During the one-month experiment, the status of the students' English listening comprehension and vocabulary acquisition was evaluated on a weekly basis. From the e...

The purpose of this study was to investigate the impact of three keyword caption modes on the listening comprehension of Arab learners of English as a foreign language (N = 90) while viewing authentic video clips. The keyword caption modes contained approximately 10%, 30% or 50% of the words in the video scripts. The participants watched three different video clips from three science videos, ea...

mesh headings,heading words, keyword]13 (accura$ or reliab$ or prevalen$ or yield$).mp. [mp=title, original title, abstract, meshheadings, heading words,keyword]14 10 or 11 or 12 or 1315 exp Mass Screening/16 (screen$ or diagnos$ or test$ or detect$).mp. [mp=title, original title, abstract, meshheadings, heading words, keyword]17 15 or 1618 9 and 17 Datab...

Fig. S1. Schematic drawings of the experimental setups. Fig. S2. Image processing performed to obtain LCM-DIM images presented in Fig. 1A. Caption of Movie S1. The coalescence of round liquid-like droplets (α-QLLs) on prism and high-index faces. Caption of Movie S2. The appearance of thin liquid-like layers (β-QLLs) beneath αQLLs. Caption of Movie S3. The coalescence of β-QLLs on prism and high...

même si dans un roman, on le sait depuis longtemps, le narrateur n’est pas une entité humaine et psychologique mais plutôt une construction poético-sémiotique, il a pourtant une « figure » et une « zone » dans laquelle elle évolue et agit. se fondant sur la théorie du point de vue d’alain rabatel et la théorie aristotélicienne de l’ethos, nous analysons les modalités de la présence du narrateur...