View Video Research Articles

Human or Humanoid Animated Pedagogical Avatars in Video Lectures: The Impact of the Knowledge Type on Learning Outcomes

This study examines the impacts of educational videos with human or humanoid animated pedagogical avatars (APAs) on cognitive, affective, and social learning outcomes in declarative and procedural knowledge contexts. A 2 × 2 factorial design was used, considering instructor type (human vs. humanoid APA) and knowledge type (declarative vs. procedural). The study involved 139 university students with no prior knowledge of the video content and was conducted in a human-computer interaction lab using an eye-tracking device. There were no significant differences in learning achievement, perceived lecture engagement, emotion, satisfaction, and social presence between the human and humanoid APA. However, learners showed greater visual interest in the human instructor, which influenced their interest in the learning content. Viewers of declarative knowledge videos showed more visual interest and achieved higher learning scores, while procedural videos increased lecture engagement, emotion, satisfaction, and social presence. The study suggests that humanoid APAs can effectively replace human instructors in educational videos and emphasizes the importance of considering knowledge types in video design to impact learning outcomes.

The Interactive Effects of Video Type and Viewing Device on Pre-service Teachers’ Learning Performance

Classroom video use in teacher education is rising, but the interplay of video type (2D, panoramic) and viewing device (PC, HMD) on learning remains unexplored. This study examined the interactive effects of video type (2D, panoramic) and viewing device (PC, Head-Mounted Display [HMD]) on pre-service teachers’ learning performance. 143 pre-service teachers from a China university participated in a 2 × 2 randomized controlled experiment, viewing a 10-minute instructional video. Results indicate that panoramic video with HMD yielded the best learning outcomes, enhancing academic achievement, attention, presence, and immersion. Conversely, 2D video on HMD increased cognitive load. Interestingly, while PC-viewed panoramic video maximized attention, it also generated the highest cognitive load. Focus group interviews revealed the potential of immersive learning alongside challenges like motion sickness and navigation difficulties. This research provides empirical evidence for the interaction of video type and viewing devices in teacher education, informing the design of effective video-based learning.

Detection and Statistics System of Pavement Distresses Based on Street View Videos

Detection and Statistics System of Pavement Distresses Based on Street View Videos

Innovative ongoing support within a multifaceted computational thinking professional learning program improves teachers’ self-efficacy and classroom practices

In-service elementary school teachers need training to teach computational thinking (CT). A common model for CT professional learning programs includes an initial workshop followed by ongoing support. However, little research has examined the efficacy of ongoing support, the most effective types, and its impact on classroom practices. The current study details a multifaceted CT professional learning program with innovative ongoing support to help elementary school teachers implement the Code.org curriculum. It also empirically investigated the impact of the support on teachers' self-efficacy in coding and teaching CT, the effectiveness of each component of the support, and to what extent it affected teachers' classroom practices. Our ongoing support included an additional workshop, grade-level professional learning community (PLC) sessions, instructional videos, and teacher leaders’ support. Data included a pre-survey, four post-surveys, lesson observations, a focus group interview, tracking the number of video views, and surveys from PLC participants and teacher leaders. Results indicated that the ongoing support significantly increased teachers’ self-efficacy in coding and teaching CT. Regarding the effectiveness of the various components of the ongoing support, the additional workshop significantly increased teachers’ self-efficacy in coding and teaching CT. Teachers valued each component of the support, and this support improved teachers’ classroom practices.

From likes to luggage: The role of social media content in attracting tourists

The tourism industry has become increasingly information-driven where tourists need reliable information to make their travel decisions. Social media content creators have emerged as prominent figures in disseminating authentic and credible information about tourist destinations worldwide. However, scant attention has been paid to explore the role of social media content in influencing tourist inflows and tourist engagement with tourism-related content on social media. To fill this gap, the present study utilized three social media analytic tools: VidIQ, Tubebuddy, and Social Blade, to collect data pertaining to tourism-related content across six tourist destinations in the Arabian Gulf region. The results indicate that these six tourist destinations collectively attracted 47 million visitors in a single year while garnering 158 million views on YouTube. Our findings indicate that tourists’ consumption of tourism-related content on YouTube directly relates to the number of tourists visiting a destination. In addition, the present study also finds that destination popularity leads to higher number of video views on YouTube as well as higher numbers of tourists to the destination. Furthermore, the current study delineates the role of content creator characteristics in influencing user engagement with tourism-related content. The present study offers valuable insights for destination management organizations, tourism operators, tourism industry stakeholders, content creators, social media marketing managers and destination brand managers.

Video as World Mandala: Nakajima Ko’s Animaker, Aniputer, and Cosmic Video

Abstract Early video processing tools present a compelling perspective through which to reconsider the integrated practice of Tokyo-based early video artist Nakajima Ko. The author examines Nakajima's practice through his concepts of animation, the development of his Animaker (1979) and Aniputer (1982) the ways he sees video connected to Esoteric Buddhist thought. Returning then to his 1986 video Mt. Fuji, the author illustrates how it is representative of the hybrid analog-digital view of video that leads Nakajima to posit it as a cosmic medium circa 1986.

Analysis of Student Activities Based on Log Files in E-Learning Using Clustering Algorithm

This study was conducted to analyse log data generated by e-learning platforms such as Moodle or similar platforms. The main objective of this research is to identify patterns and insights that can help improve the student learning experience and the efficiency of platform management. This study identifies the most effective clustering algorithms for grouping students based on their behaviour and achievements in the e-learning environment, namely K-means and K-Medoids. The methods used to determine the optimal number of clusters are the Silhouette Coefficient method and the Elbow method, utilizing both methods to determine the best clustering algorithm results. Based on the analysis using K-means and K-Medoids clustering methods on the log data of the programming algorithm course, the total number of action logs over a one-month period is 95,461, with an average of 892 action logs per participant. The distribution of action logs based on class (A, B, and C) shows variation in the number and average of action logs per class. Types of activities such as assignments, forums, video views, and course views have different average frequencies, with video views being the most frequently visited activity. Pearson correlations between activity types show strong relationships between activities, with the highest correlation between visits and course views. The optimal number of clusters based on the Elbow and K-Medoids methods is three clusters. Cluster 3 in K-Means has the best performance with the smallest DBI value (0.424) and the smallest centroid distance (21,168.534).

INFÂNCIA, PRODUÇÕES DE CONTEÚDOS DIGITAIS E RELAÇÕES TRANSMÍDIA ENTRE O YOUTUBE E O TIKTOK

This article aims to investigate possible transmedia relationships between the YouTube and TikTok platforms within the context of content targeted at children, created by digital creators. Using an exploratory approach, three digital content creators, namely Luccas Neto, Manual do Mundo, and Enaldinho, were chosen as they operate on both platforms and produce content for children. The Socialblade platform provided data on the number of subscribers and the number of views of the most-watched videos, which were used as parameters in this study. Based on the results, we identified that content creators who focus on children's stories are more inclined to incorporate and utilize transmedia concepts and characteristics, as they produce and distribute their content across both platforms, with the intention of engaging children.

Bitcoin returns and YouTube news: a behavioural time series analysis

ABSTRACT This study investigates whether investor’s sentiment and attention information collected via YouTube can improve bitcoin return forecasts. Accordingly, we collected daily data over the period 2017–2023, covering calm and turbulent periods marked by different types and episodes of emotions. Unlike previous studies, we used YouTube videos to propose two sentiment proxies: investor attention to YouTube (daily number of YouTube video views) and investor sentiment on YouTube (number of positive and negative videos on YouTube). Interestingly, we break down both attention and sentiment per subject. Econometrically, we assess lead-lag effects between sentiment/attention and bitcoin return using causality tests and Vector Auto-regressive (VAR) model. We also evaluate the forecasting power of YouTube attention/sentiment data using a deep learning LSTM model. Our study shows two main results. First, we find lead-lag effects between bitcoin returns and per subject investor’s attention and sentiment proxies. Second, we show that our deep learning LSTM model relying on the information provided by attention and sentiment supplants benchmark Buy and Hold Strategy to forecast future bitcoin returns.

Multimodal Cross-lingual Summarization for Videos: A Revisit in Knowledge Distillation Induced Triple-stage Training Method.

Multimodal summarization (MS) for videos aims to generate summaries from multi-source information (e.g., video and text transcript), and this technique has made promising progress recently. However, existing works are limited to monolingual video scenarios, overlooking the demands of non-native language video viewers to understand cross-lingual videos in practical applications. It stimulates us to introduce multimodal cross-lingual summarization for videos (MCLS), which aims at generating cross-lingual summarization from multimodal input of videos. Considering the challenge of high annotation cost and resource constraints in MCLS, we propose a knowledge distillation (KD) induced triple-stage training method to assist MCLS by transferring knowledge from abundant monolingual MS data to those data with insufficient volumes. In the triple-stage training method, a video-guided dual fusion network (VDF) is designed as the backbone network to integrate multimodal and cross-lingual information through different fusion strategies of encoder and decoder; what's more, we propose two cross-lingual knowledge distillation strategies: adaptive pooling distillation and language-adaptive warping distillation (LAWD). These strategies are tailored for distillation objects (i.e., encoder-level and vocab-level KD) to facilitate effective knowledge transfer across cross-lingual sequences of varying lengths between MS and MCLS models. Specifically, to tackle the challenge of unequal length of parallel cross-language sequences in KD, our proposed LAWD can directly conduct cross-language distillation while keeping the language feature shape unchanged to reduce potential information loss. We meticulously annotated the How2-MCLS dataset based on the How2 dataset to simulate the MCLS scenario. The experimental results show that the proposed method achieves competitive performance compared to strong baselines, and can bring substantial performance improvements to MCLS models by transferring knowledge from the MS model.

Lessons learned from naturalistic driving data processing in a secure data enclave: Preliminary discoveries from analyzing dash camera videos

This paper provides preliminary insights on the challenges of processing Strategic Highway Research Program 2 (SHRP2) Naturalistic Driving Study (NDS) videos and data, particularly those with Personally Identifiable Information (PII). Insights and lessons learned are presented from a study designed to evaluate the effectiveness of High Visibility Crosswalks (HVCs). Over a one-month period, 15,379 videos were processed in the secure data enclave of Virginia Tech Transportation Institute (VTTI). As these videos are not available outside of the secure data enclave due to PII restrictions, researchers visiting the secure data enclave for the first time may face several challenges: navigating the software interface; identifying the video views and frames of interest; and identifying and extracting information of interest from the video views, etc. These challenges, the procedures followed to address them, and the process for identifying and classifying distracted driving behaviors are discussed. Lastly, hypothesis tests are conducted to investigate distracted driving behavior, with the results revealing that HVCs have the potential to make drivers more cautious in their proximity. The information presented in this paper is expected to aid researchers who intend to utilize SHRP2 NDS or similar videos for future research, to preemptively plan for the video processing phase.

High-Throughput Deep Learning Detection of Mitral Regurgitation.

Diagnosis of mitral regurgitation (MR) requires careful evaluation by echocardiography with Doppler imaging. This study presents the development and validation of a fully automated deep learning pipeline for identifying apical 4-chamber view videos with color Doppler echocardiography and detecting clinically significant (moderate or severe) MR from transthoracic echocardiograms. A total of 58 614 transthoracic echocardiograms (2 587 538 videos) from Cedars-Sinai Medical Center were used to develop and test an automated pipeline to identify apical 4-chamber view videos with color Doppler across the mitral valve and then assess MR severity. The model was tested internally on a test set of 1800 studies (80 833 videos) from Cedars-Sinai Medical Center and externally evaluated in a geographically distinct cohort of 915 studies (46 890 videos) from Stanford Healthcare. In the held-out Cedars-Sinai Medical Center test set, the view classifier demonstrated an area under the curve (AUC) of 0.998 (0.998-0.999) and correctly identified 3452 of 3539 echocardiography videos as having color Doppler information across the mitral valve (sensitivity of 0.975 [0.968-0.982] and specificity of 0.999 [0.999-0.999] compared with manually curated videos). In the external test cohort from Stanford Healthcare, the view classifier correctly identified 1051 of 1055 manually curated videos with color Doppler information across the mitral valve (sensitivity of 0.996 [0.990-1.000] and specificity of 0.999 [0.999-0.999]). In the Cedars-Sinai Medical Center test cohort, MR moderate or greater in severity was detected with an AUC of 0.916 (0.899-0.932) and severe MR was detected with an AUC of 0.934 (0.913-0.953). In the Stanford Healthcare test cohort, the model detected MR moderate or greater in severity with an AUC of 0.951 (0.924-0.973) and severe MR with an AUC of 0.969 (0.946-0.987). In this study, a novel automated pipeline for identifying clinically significant MR from full transthoracic echocardiography studies demonstrated excellent performance across large numbers of studies and across multiple institutions. Such an approach has the potential for automated screening and surveillance of MR.

Towards the use of virtual reality prototypes in architecture to collect user experiences: An assessment of the comparability of patient experiences in a virtual and a real ambulatory pathway

Virtual Reality (VR) enables the low-cost production of realistic prototypes of buildings at early stages of architectural projects. Such prototypes may be used to gather the experiences of future users and iterate early on in the design. However, it is essential to evaluate whether what is experienced within such VR prototypes corresponds to what will be experienced in reality. Here, we use an innovative method to compare the experiences of patients in a real building and in a virtual environment that plays the role of a prototype that could have been created by architects during the design phase. We first designed and implemented a VR environment replicating an existing ambulatory pathway. Then, we used micro-phenomenological interviews to collect the experiences of real patients in the VR environment (n=8), along with VR traces and first-person point of view videos, and in the real ambulatory pathway (n=8). We modeled and normalized the experiences, and compared them systematically. Results suggest that patients live comparable experiences along various experiential dimensions such as thought, emotion, sensation, social and sensory perceptions, and that VR prototypes may be adequate to assess issues with architectural design. This work opens unique perspectives towards involving patients in User-Centered Design in architecture, though challenges lie ahead in how to design VR prototypes from early blueprints of architects.

Severity of inattention symptoms, experiences of being bullied, and school anxiety as mediators in the association between excessive short-form video viewing and school refusal behaviors in adolescents.

Recent years have seen an increase in school refusal behavior among adolescents, potentially due to factors like excessive short-form video viewing, bullying, and school anxiety. Limited research has investigated how these factors contribute to school refusal behavior. This study used random forest regression, path analysis, and network analysis to identify key variables and pathways leading to school refusal behavior. In this cross-sectional questionnaire-based study, 2,056 (996 male, 1,060 female, mean age: 14.79 ± 1.24 years) middle and senior high school students were asked to complete the School Refusal Behavior Assessment questionnaire to assess school refusal behavior features, the Excessive Short-Form Video Viewing Scale as well as self-reported viewing times during leisure days to assess excessive short-form video viewing, the SNAP-IV Rating Scale to assess the severity of inattention symptoms, and the self-administered questionnaires to assess experiences of being bullied and school anxiety. The prevalence of school refusal behavior in the surveyed adolescents was found to be 31.9% [95% confidence interval (CI): 29.8-33.9%]. In terms of significance, the severity of inattention symptoms exhibited the greatest predictive power, while excessive short-form video viewing accounted for the most variance. Path analysis revealed that excessive short-form video viewing not only directly affects school refusal behavior features but also does so indirectly through severity of inattention symptoms and school anxiety. Key bridge factors in this pathway include intense fear and anxiety associated with school attendance, manifesting as somatic symptoms and avoidance behaviors. The findings indicate that not only does excessive short-form video viewing directly influence school refusal behavior features in adolescents, but it also indirectly impacts these features through mechanisms involving severity of inattention symptoms and school anxiety. The bridge factors highlight potential targets for interventions among the SRB features and predictors.

YouTube videos on lymphedema as an information source for Spanish speaking breast cancer survivors

BackgroundBreast cancer-related lymphedema in the upper limb remains one of the most distressful complications of breast cancer treatment. YouTube is considered a potential digital resource for population health and decision making. However, access to inadequate information or misinformation could have undesirable impacts. This cross-sectional study aimed to evaluate the reliability, quality and content of YouTube videos on lymphedema as an information source for Spanish-speaking breast cancer survivors.MethodsA search of YouTube was conducted in January 2023 using the key words “breast cancer lymphedema” and “lymphedema arm breast cancer.” Reliability and quality of the videos were evaluated using the Discern tool, content, source of production, number of likes, comments, views, duration, Video Power Index, likes ratio, view ratio and age on the platform.ResultsAmongst the 300 Spanish language videos identified on YouTube, 35 were selected for analysis based on the inclusion and exclusion criteria. Of the 35 selected videos, 82.9% (n = 29) were developed by healthcare or academic professionals and 17.1% (n = 9) by others. Reliability (p < 0.017) and quality (p < 0.03) were higher in the videos made by professionals. The Discern total score (r = 0.476; p = 0.004), reliability (r = 0.472; p = 0.004) and quality (r = 0.469; p = 0.004) were positively correlated with the duration of the videos.ConclusionsOur findings provide a strong rationale for educating breast cancer survivors seeking lymphedema information to select videos made by healthcare or academic professionals. Standardised evaluation prior to video publication is needed to ensure that the end-users receive accurate and quality information from YouTube.

Clicks for money: Predicting video views through a sentiment analysis of titles and thumbnails

Videos dominate online marketing and social media content, and therefore making a video thumbnail highly clickable is a critical yet challenging task. Intense competition among vloggers and channel owners for advertising revenue and sponsorship has led to the increasing use of clickbait-style thumbnails and titles with exaggerated sentiments to boost click-through rates. Drawing on cognitive schema theory, we analyzed the sentiments in video titles, thumbnails, and captions to examine their effects on video views. Our analysis of 16,215 YouTube video covers reveals that strong sentiments in thumbnails, whether positive or negative, lead to more views, whereas strong sentiments in captions have the opposite effect. Video titles with positive rather than negative sentiments generate more views. Sentiment markers such as punctuation, all capitals, and emoticons all contribute to more clicks. These findings highlight that the use of sentiment in video covers can have both a positive and a negative effect, which has significant implications for video marketing.

Optical coherence tomography homography for detection of retinal displacement: a validation study.

Retinal displacement following rhegmatogenous retinal detachment (RRD) has been associated with inferior functional outcomes. Recent evidence using an overlay technique suggests that fundus-autofluorescence underestimates post-RRD repair retinal displacement. This study aims to validate the overlay technique in normal eyes and to determine its sensitivity and specificity at detecting retinal displacement. We conducted a retrospective case series involving 66 normal eyes, each with at least two separate infrared (IR) images at different time points. Overlay of the two images was based on manual marking of choroidal and optic nerve head (ONH) landmarks. For each set of two IR images, computer code for homography generated two outputs, flipping view video and an overlay picture. First, validation of choroidal/ONH alignment was performed using the flipping view video to ensure accurate manual markings. Then, two different masked graders (AB + IM) evaluated the overlays for presence of retinal displacement. 16 control eyes following RRD repair with detected retinal displacement on FAF imaging assessed sensitivity and specificity of the technique. 94% of overlays were found to be well aligned (62/66). 11 cases exhibited errors on flipping view analysis (choroidal/ONH misalignment). Those 11 cases had a significantly higher rate of retinal displacement (false positives) compared to cases without errors (8/11,72% Vs 54/55,98%,P = 0.001). Sensitivity and specificity of the overlay technique for detecting retinal displacement considering only adequate flipping view cases (n = 55) were calculated as 100% and 98%, respectively. IR overlay emerges as a reliable and valid method for detecting retinal displacement, exhibiting excellent sensitivity and specificity.

Social Media Programs for Outreach and Recruitment Supporting Aging and Alzheimer Disease and Related Dementias Research: Longitudinal Descriptive Study.

Social media may be a useful method for research centers to deliver health messages, increase their visibility in the local community, and recruit study participants. Sharing examples of social media-based community outreach and educational programs, and evaluating their outcomes in this setting, is important for understanding whether these efforts have a measurable impact. The aim of this study is to describe one center's social media activities for community education on topics related to aging, memory loss, and Alzheimer disease and related dementias, and provide metrics related to recruitment into clinical research studies. Several social media platforms were used, including Facebook, X (formerly Twitter), and YouTube. Objective assessments quantified monthly, based on each platform's native dashboard, included the number of followers, number of posts, post reach and engagement, post impressions, and video views. The number of participants volunteering for research during this period was additionally tracked using a secure database. Educational material posted to social media most frequently included content developed by center staff, content from partner organizations, and news articles or resources featuring center researchers. Multiple educational programs were developed, including social media series, web-based talks, Twitter chats, and webinars. In more recent years, Facebook content was occasionally boosted to increase visibility in the local geographical region. Up to 4 years of page metrics demonstrated continuing growth in reaching social media audiences, as indicated by increases over time in the numbers of likes or followers on Facebook and X/Twitter and views of YouTube videos (growth trajectories). While Facebook reach and X/Twitter impression rates were reasonable, Facebook engagement rates were more modest. Months that included boosted Facebook posts resulted in a greater change in page followers and page likes, and higher reach and engagement rates (all P≤.002). Recruitment of participants into center-affiliated research studies increased during this time frame, particularly in response to boosted Facebook posts. These data demonstrate that social media activities can provide meaningful community educational opportunities focused on Alzheimer disease and related dementias and have a measurable impact on the recruitment of participants into research studies. Additionally, this study highlights the importance of tracking outreach program outcomes for evaluating return on investment.

How Does Short Video Viewing Influence Young Children’s Everyday Language Practices? A Case Study of China

How Does Short Video Viewing Influence Young Children’s Everyday Language Practices? A Case Study of China

Deep Learning Phenotyping of Tricuspid Regurgitation for Automated High Throughput Assessment of Transthoracic Echocardiography.

Diagnosis of tricuspid regurgitation (TR) requires careful expert evaluation. This study developed an automated deep learning pipeline for assessing TR from transthoracic echocardiography. An automated deep learning workflow was developed using 47,312 studies (2,079,898 videos) from Cedars-Sinai Medical Center (CSMC) between 2011 and 2021. The pipeline was tested on a temporally distinct test set of 2,462 studies (108,138 videos) obtained in 2022 at CSMC and a geographically distinct cohort of 5,549 studies (278,377 videos) from Stanford Healthcare (SHC). In the CSMC test dataset, the view classifier demonstrated an AUC of 1.000 (0.999 - 1.000) and identified at least one A4C video with colour Doppler across the tricuspid valve in 2,410 of 2,462 studies with a sensitivity of 0.975 (0.968-0.982) and a specificity of 1.000 (1.00-1.000). In the CSMC test cohort, moderate-or-severe TR was detected with an AUC of 0.928 (0.913 - 0.943) and severe TR was detected with an AUC of 0.956 (0.940 - 0.969). In the SHC cohort, the view classifier correctly identified at least one TR colour Doppler video in 5,268 of the 5,549 studies, resulting in an AUC of 0.999 (0.998 - 0.999), a sensitivity of 0.949 (0.944 - 0.955) and specificity of 0.999 (0.999 - 0.999). The AI model detected moderate-or-severe TR with an AUC of 0.951 (0.938 - 0.962) and severe TR with an AUC of 0.980 (0.966 - 0.988). We developed an automated pipeline to identify clinically significant TR with excellent performance. This approach carries potential for automated TR detection and stratification for surveillance and screening. Can an automated deep learning model assess tricuspid regurgitation severity from echocardiography? We developed and validated an automated tricuspid regurgitation detection algorithm pipeline across two healthcare systems with high volume echocardiography labs. The algorithm correctly identifies apical-4-chamber view videos with colour Doppler across the tricuspid valve and grades clinically significant TR with strong agreement to expert clinical readers. A deep learning pipeline could automate TR screening, facilitating reproducible accurate assessment of TR severity, allowing rapid triage or re-review and expand access in low-resource or primary care settings.