Search Results (62)

This study examined the content characteristics and user responses of athlete-run sports YouTube channels, providing empirical insights for content production strategies and contributing to the development of athlete-run sports YouTube channels. Content analysis was conducted on 3306 videos posted on 20 popular YouTube channels of South Korean athletes from 1 January 2020 to 31 December 2021. The formal characteristics analyzed included video length, the presence of foreign language subtitles, paid advertisements, and information sources. The content characteristics examined were the types of sports events, main content themes, and whether the content matched the athlete’s sport. Results revealed significant differences in content characteristics and user responses based on whether the athletes were active or retired. This study’s distinctive contribution lies in highlighting the evolving role of athletes as content creators and providing strategic implications for enhancing the competitiveness of athlete-run sports YouTube channels. Future research should consider a broader range of sports YouTubers and a wider variety of YouTube channels to gain comprehensive insights into the sports content ecosystem on this platform. Full article

Speech comprehension can be challenging due to multiple factors, causing inconvenience for both the speaker and the listener. In such situations, using a humanoid robot, Pepper, can be beneficial as it can display the corresponding text on its screen. However, prior to that, it is essential to carefully assess the accuracy of the audio recordings captured by Pepper. Therefore, in this study, an experiment is conducted with eight participants with the primary objective of examining Pepper’s speech recognition system with the help of audio features such as Mel-Frequency Cepstral Coefficients, spectral centroid, spectral flatness, the Zero-Crossing Rate, pitch, and energy. Furthermore, the K-means algorithm was employed to create clusters based on these features with the aim of selecting the most suitable cluster with the help of the speech-to-text conversion tool Whisper. The selection of the best cluster is accomplished by finding the maximum accuracy data points lying in a cluster. A criterion of discarding data points with values of WER above 0.3 is imposed to achieve this. The findings of this study suggest that a distance of up to one meter from the humanoid robot Pepper is suitable for capturing the best speech recordings. In contrast, age and gender do not influence the accuracy of recorded speech. The proposed system will provide a significant strength in settings where subtitles are required to improve the comprehension of spoken statements. Full article

Subtitles play a crucial role in facilitating the understanding of visual content when watching films and television programs. In this study, we propose a method for presenting subtitles in a way that considers cognitive load when viewing video content in a non-native language. Subtitles are generally displayed at the bottom of the screen, which causes frequent eye focus switching between subtitles and video, increasing the cognitive load. In our proposed method, we focused on the position, display time, and amount of information contained in the subtitles to reduce the cognitive load and to avoid disturbing the viewer’s concentration. We conducted two experiments to investigate the effects of our proposed subtitle method on gaze distribution, comprehension, and cognitive load during English-language video viewing. Twelve non-native English-speaking subjects participated in the first experiment. The results show that participants’ gazes were more focused around the center of the screen when using our proposed subtitles compared to regular subtitles. Comprehension levels recorded using LightSub were similar, but slightly inferior to those recorded using regular subtitles. However, it was confirmed that most of the participants were viewing the video with a higher cognitive load using the proposed subtitle method. In the second experiment, we investigated subtitles considering connected speech form in English with 18 non-native English speakers. The results revealed that the proposed method, considering connected speech form, demonstrated an improvement in cognitive load during video viewing but it remained higher than that of regular subtitles. Full article

Reception studies in audiovisual translation seek to explore how translation choices affect the audience’s comprehension, emotional engagement, enjoyment, and overall viewing experience of audiovisual materials. This study focuses on the subtitling product and analyzes the acceptability of swear words translated through different stimuli: subtitles with softened, maintained, and intensified swearing, along with standard Netflix subtitles (control). Employing a multi-method approach, the study collected data through a survey, using questionnaires with a Likert scale and interviews, following the user-centered translation model to understand how participants receive and perceive swear words in subtitling. The results indicate that the control group had the highest acceptability of the participants, while the group with softened swear words presented the lowest acceptability rate. The analysis shows that participants across all groups reported that discomfort does not arise from reading the swear word in the subtitle but from perceiving a deliberate change in its offensive load—usually softened. The findings demonstrate that this change can lead to a breach of the contract of illusion in subtitling, as participants are exposed to the original dialogue and the translated subtitle simultaneously. In conclusion, when perceived, the change in the offensive load can redirect the viewer’s focus from the video to the subtitles, negatively affecting the enjoyment of the audiovisual experience. Full article

Learning a Foreign Language (FL) beyond the classroom has become common practice thanks to advances in technology and the use of English as a Lingua Franca. This study explores the types and amount of out-of-school informal exposure to English that Spanish secondary school students typically receive in their daily lives. Informed by recent literature on the influence of extramural activities on FL proficiency, the second aim of this study is to investigate the potential relationship between out-of-school exposure and academic performance, as measured by English school grades. Data were obtained from a questionnaire answered by secondary school students aged 12–16 (N = 2015) regarding the different types and amounts of activities they perform in English outside school. Findings revealed that teenage learners were most frequently exposed to English through audiovisual input. Social media interaction, along with reading and writing (with or without digital support), were closely associated with their English marks. Other popular activities, such as listening to music or playing video games, were not found to be related to proficiency or even showed a negative correlation with it, while less popular activities, such as watching subtitled movies and series, could have greater potential for language learning. This study contributes to the understanding of informal practices in FL learning settings and provides insights that can help bridge interactive language practices and formal curriculum to create holistic learning experiences for language learners. Full article

In the digital transformation era, video media libraries’ untapped potential is immense, restricted primarily by their non-machine-readable nature and basic search functionalities limited to standard metadata. This study presents a novel multimodal methodology that utilizes advances in artificial intelligence, including neural networks, computer vision, and natural language processing, to extract and geocode geospatial references from videos. Leveraging the geospatial information from videos enables semantic searches, enhances search relevance, and allows for targeted advertising, particularly on mobile platforms. The methodology involves a comprehensive process, including data acquisition from ARD Mediathek, image and text analysis using advanced machine learning models, and audio and subtitle processing with state-of-the-art linguistic models. Despite challenges like model interpretability and the complexity of geospatial data extraction, this study’s findings indicate significant potential for advancing the precision of spatial data analysis within video content, promising to enrich media libraries with more navigable, contextually rich content. This advancement has implications for user engagement, targeted services, and broader urban planning and cultural heritage applications. Full article

To create a more immersive experience, electronic content developers utilize hardware solutions that not only display images and produce sounds but also manipulate the viewer’s real environment. These devices can control visual effects like lighting variations and fog, emit scents, simulate liquid effects, and provide vibration or locomotion sensations, such as moving the viewer’s chair. The goal is to emulate additional sensations for the viewers and engender the belief that they are truly present within the virtual environment. These devices are typically found in specially designed cinemas referred to as xD cinemas, such as 4D, 5D, 9D, etc., where each effect is treated as an additional dimension, enhancing the overall experience. Currently, all of these effects are triggered by timers. The system determines which effect to play based on timers. This approach is problematic, for it requires programming each device for each movie. In this research, we address this problem by introducing the idea of Special Effect Tags (SETs) that can be added in the subtitle files. The SETs aim to serve as a standard that will allow the various devices to know when each artificial phenomenon should be triggered. They are generic and can support infinite artificial phenomena, also known as dimensions. This paper introduces the idea of a common special effect framework and a generic architecture of a special effects player that is independent of any specific hardware solutions. Full article

The InnoDAT project is framed within the TRADILEX Project, which is aimed at demonstrating the applicability of Audiovisual Translation (AVT) for teaching and learning languages. TRADILEX is an ongoing project presented at a state-funded competitive call and supported by the Spanish Government. This article is aimed at presenting InnoDAT, an innovative project based in the use of AVT for teaching and learning languages through: Subtitling for the Deaf and Hard-of-Hearing (SDH). It has been designed to learn Spanish as a Second Foreign Language in an Italian Higher Education context according to a B2 CEFR level. The methodology used was developed by researchers of TRADILEX. Six tailormade Learning Units (LU), based on the SDH mode, were designed and implemented among participants (N = 97). Authentic materials and cultural matters were also used and adapted according to the B2 level. The results show a clear improvement in the process of teaching and learning languages, knowledge of the culture and traditions of the target language, and the consciousness of accessibility among the participants. The authors compare this innovative research with former research The InnoDAT project validates the applicability of the didactic audiovisual translation (DAT) as a means for learning languages and cultures within digital educational settings and how languages and cultures are intricately connected. Moreover, not only cultural issues but also accessibility were paramount in this research. Finally, motivation, autonomous and meaningful learning, communicative language competence, and digital competence were also nurtured by means of the InnoDAT project. Full article

The explosive growth of online short videos has brought great challenges to the efficient management of video content classification, retrieval, and recommendation. Video features for video management can be extracted from video image frames by various algorithms, and they have been proven to be effective in the video classification of sensor systems. However, frame-by-frame processing of video image frames not only requires huge computing power, but also classification algorithms based on a single modality of video features cannot meet the accuracy requirements in specific scenarios. In response to these concerns, we introduce a short video categorization architecture centered around cross-modal fusion in visual sensor systems which jointly utilizes video features and text features to classify short videos, avoiding processing a large number of image frames during classification. Firstly, the image space is extended to three-dimensional space–time by a self-attention mechanism, and a series of patches are extracted from a single image frame. Each patch is linearly mapped into the embedding layer of the Timesformer network and augmented with positional information to extract video features. Second, the text features of subtitles are extracted through the bidirectional encoder representation from the Transformers (BERT) pre-training model. Finally, cross-modal fusion is performed based on the extracted video and text features, resulting in improved accuracy for short video classification tasks. The outcomes of our experiments showcase a substantial superiority of our introduced classification framework compared to alternative baseline video classification methodologies. This framework can be applied in sensor systems for potential video classification. Full article

Delving into the realm of massive open online courses (MOOCs), this investigation scrutinizes their role in enriching the professional language proficiency of economics undergraduates amidst a global coronavirus crisis. The research pivots on assessing the efficacy of MOOCs within higher learning curriculums while shedding light on the benefits and drawbacks of such an approach. The methodologies employed encompassed the analysis and interpretation of data derived from student polls and statistical analysis of yielded outcomes. The findings shed light on MOOCs as a potent instrument for honing language prowess and enriching the professional linguistic expertise of budding economists. The language experiment was participated in by a group of 34 students and their responses have been collated and interpreted as proposed guidelines for future reference. The participants notably appreciated the video-lecture format and exposure to native speakers’ elocution. However, the study did not shy away from spotlighting certain limitations such as complexities in platform navigation and the difficulties of watching videos without subtitles. To optimize MOOC utilization, a series of recommendations were drawn up. These included offering choices to students, simplifying platform navigation, extending support during challenges, and bolstering speaking abilities. This exploration holds valuable insights not just for educators and students but also for academic institutions at large, offering hands-on data regarding the perks and pitfalls of employing MOOCs to cultivate professional foreign language competence within the field of economic education. Full article

The discipline of automatic image captioning represents an integration of two pivotal branches of artificial intelligence, namely computer vision (CV) and natural language processing (NLP). The principal functionality of this technology lies in transmuting the extracted visual features into semantic information of a higher order. The bidirectional long short-term memory (Bi-LSTM) has garnered wide acceptance in executing image captioning tasks. Of late, scholarly attention has been focused on modifying suitable models for innovative and precise subtitle captions, although tuning the parameters of the model does not invariably yield optimal outcomes. Given this, the current research proposes a model that effectively employs the bidirectional LSTM and attention mechanism (Bi-LS-AttM) for image captioning endeavors. This model exploits the contextual comprehension from both anterior and posterior aspects of the input data, synergistically with the attention mechanism, thereby augmenting the precision of visual language interpretation. The distinctiveness of this research is embodied in its incorporation of Bi-LSTM and the attention mechanism to engender sentences that are both structurally innovative and accurately reflective of the image content. To enhance temporal efficiency and accuracy, this study substitutes convolutional neural networks (CNNs) with fast region-based convolutional networks (Fast RCNNs). Additionally, it refines the process of generation and evaluation of common space, thus fostering improved efficiency. Our model was tested for its performance on Flickr30k and MSCOCO datasets (80 object categories). Comparative analyses of performance metrics reveal that our model, leveraging the Bi-LS-AttM, surpasses unidirectional and Bi-LSTM models. When applied to caption generation and image-sentence retrieval tasks, our model manifests time economies of approximately 36.5% and 26.3% vis-a-vis the Bi-LSTM model and the deep Bi-LSTM model, respectively. Full article

Multilingualism and multiculturalism are verbally and visually recurrent in the sitcom Family Guy (1999-in production) through a combination of a main language of communication (L1) and other languages (L3) in the source language (SL) or source text (ST). The use of L3 is tantamount to tokenism and stereotyping characters, especially those whose recurrence is incidental and part of jokes. This paper compares two versions of the episode “Road to Rhode Island” (American and Spanish DVDs) and addresses a scene to analyze the linguistic challenges and lexical choices in dubbing and subtitling L1 and L3 in two geographical varieties of Spanish: Latin American Spanish and Peninsular Spanish. In this regard, this study focuses on the role and function of L3 in translation, the techniques to represent L3 in translation, L1 and L3 translation techniques, and which techniques are used in translation. Overall, this paper explores how the Spanish DVD adds a new L3 in the target text (TT) to maintain its original function in subtitling and dubbing, and the differences in the American DVD: L3^TT omission in subtitling and L3^TT change of function and meaning in dubbing, which ultimately accentuates linguistic and cultural misrepresentation and stereotypes. Full article

Linguistic diversity is present in many audiovisual productions and has given rise to fruitful research on translation of multilingualism and language variation. Monsoon Wedding (Mira Nair, 2001) is a prototypical film for translation analysis, since multilingualism is a recurrent feature, as the film dialogue combines English (L1) with Hindi and Punjabi (L3), which creates an effect of code-switching. This article analyses how the multilingualism and the cultural elements present in the source text (ST) have been transferred to the Spanish translated text (TT) La boda del monzón. The results show that in the Spanish dubbed and subtitled versions, few Indian cultural elements are left, and little language variation is preserved. Thus, L3 does not play a central role as it does in the source text. In the translation, only a few loan words from Hindi or Punjabi are kept, mainly from the domains of food and cooking, as well as terms of address and greetings, or words related to the wedding ceremony. The results also show that when L3 is not fully rendered in translation, otherness is still conveyed through image and music, thus (re)creating a different atmosphere for Spanish audiences. Full article

Given the vast scholarly attention paid to multilingualism on traditional media over the years, it seems timely to focus on streaming platforms. This paper sets out to identify potential norms for the rendering of multilingual occurrences in the localised content of Netflix series. It also seeks to explore whether streaming translation practices related to multilingualism differ from the consolidated norms and practices for TV and cinema content. The chosen data sample consists of the Italian dub streams of five TV Netflix-produced shows featuring multilingualism as a main characteristic. The strategies and techniques adopted in each series are singled out, quantified, and labelled according to a combination of taxonomies. These include dubbing, revoicing, subtitling, part-subtitling, diegetic interpreting, unchanged speech transfer, and no translation. A wider analysis is also carried out across all the data sample to draw patterns on a macro level. The findings reveal a strong tendency to mark and preserve multilingualism, in line with Netflix’s own policies and dubbing specifications. Transfer unchanged combined with subtitles emerges as the most recurrent strategy, while the dub-over strategy accounts for 13% of the multilingual occurrences in the data sample. Extensive neutralisation is therefore not encountered. That said, a certain degree of overlap between multilingual translation norms on Netflix and conventional Italian dubbing practices (which tend to neutralise) can still be observed. Full article

Multilingualism in audiovisual productions has substantially increased in recent years as a reflection of today’s globalised world. While the number of publications looking at the phenomenon from the perspective of audiovisual translation (AVT)—especially interlingual subtitling and dubbing—has grown considerably in the last decade, there seems to be relatively little research on the rendering of multilingualism from the perspective of accessibility modes, namely subtitling for the deaf and hard of hearing (SDH) and audio description (AD). This article aims to investigate how multilingualism is rendered for deaf and hard-of-hearing as well as blind and partially sighted audiences, focusing on SDH and AD, as well as audio subtitling (AST). The study analyses a small corpus of TV shows available on Netflix and aims to highlight how multilingualism is made accessible both in SDH and AD. The products selected for the study had to satisfy three main criteria: they had to be a recent production, include the presence of an L1 (English) and one or more third languages and offer both intralingual SDH (closed captions) and AD. The results show that, even within the context of a single streaming platform, the strategies applied to deal with multilingualism seem to vary quite significantly both in SDH and AD/AST, ranging from neutralisation to L3 visibility. Full article