Video Quality Research Articles

Analyzing dissemination, quality, and reliability of Chinese brain tumor-related short videos on TikTok and Bilibili: a cross-sectional study

Analyzing dissemination, quality, and reliability of Chinese brain tumor-related short videos on TikTok and Bilibili: a cross-sectional study

Reputation Burning: Analyzing the Impact of Brand Sponsorship on Social Influencers

The growth of the influencer marketing industry warrants an empirical examination of the effect of posting sponsored videos on influencers’ reputations. We collected a novel data set of user-generated YouTube videos created by prominent English-speaking influencers in the beauty and style category. We extracted a rich set of theory-driven video features and used DiNardo-Fortin-Lemieux reweighting to construct comparable treatment and control groups matched at the influencer-video level. The empirical analysis of the matched sample revealed a reputation-burning effect; that is, posting a sponsored video, compared with posting an equivalent organic video, cost the influencers 0.19% of their reputation (operationalized as the number of subscribers). The reputation-burning effect was stronger among the influencers with larger audiences. An analysis of likes, comments, and comment texts revealed a larger gap in audience response between sponsored and organic videos among the influencers with larger (versus smaller) audiences. The reputation-burning effect was mitigated when a high fit existed between the sponsored content and the influencer’s “usual” content and when the promoted brand was less well known. Our study empirically tested the assumption of several theoretical works. Moreover, it contributes to the literature on influencer marketing and celebrity endorsements and provides managerial implications for influencers, brands, and social media platforms. This paper was accepted by D. J. Wu, information systems. Funding: This research was supported by funding provided by Harvard Business School. Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2023.00193 .

Extract And Reconstruct: Research On the Content Of “Second Creation” Short Videos for Film Promotion

Short videos have surged in popularity, particularly on platforms like Douyin, Kuaishou, WeChat Channel, and Little Red Book in China. These brief yet engaging videos are extensively produced and consumed, making them a potent tool for promoting theatrical films. Among the various types of short videos, second-creation short videos—those that creatively reinterpret or remix original content—are particularly effective in driving platform users to theaters. This research focuses on these second-creation short videos used for film promotion. Through extensive viewing and creating these videos during an internship, the author observed notable deviations in meaning between the original films and their second-creation counterparts. To delve deeper, this study analyzes short videos related to “The Boy and the Heron”. Starting from the unique characteristics of second-creation short videos, the research identifies issues such as the narrowing of meaning that can occur in the adaptation process. By highlighting these problems, the study aims to offer insights into how audiences should interpret short videos in the context of film promotion. Finally, the research proposes recommendations for better alignment between original films and their promotional short videos, ensuring a more accurate representation of the cinematic works.

Content-Adaptive Bitrate Ladder Estimation in High-Efficiency Video Coding Utilizing Spatiotemporal Resolutions

The constant increase in multimedia Internet traffic in the form of video streaming requires new solutions for efficient video coding to save bandwidth and network resources. HTTP adaptive streaming (HAS), the most widely used solution for video streaming, allows the client to adaptively select the bitrate according to the transmission conditions. For this purpose, multiple presentations of the same video content are generated on the video server, which contains video sequences encoded at different bitrates with resolution adjustment to achieve the best Quality of Experience (QoE). This set of bitrate–resolution pairs is called a bitrate ladder. In addition to the traditional one-size-fits-all scheme for the bitrate ladder, context-aware solutions have recently been proposed that enable optimum bitrate–resolution pairs for video sequences of different complexity. However, these solutions use only spatial resolution for optimization, while the selection of the optimal combination of spatial and temporal resolution for a given bitrate has not been sufficiently investigated. This paper proposes bit-ladder optimization considering spatiotemporal features of video sequences and usage of optimal spatial and temporal resolution related to video content complexity. Optimization along two dimensions of resolution significantly increases the complexity of the problem and the approach of intensive encoding for all spatial and temporal resolutions in a wide range of bitrates, for each video sequence, is not feasible in real time. In order to reduce the level of complexity, we propose a data augmentation using a neural network (NN)-based model. To train the NN model, we used seven video sequences of different content complexity, encoded with the HEVC encoder at five different spatial resolutions (SR) up to 4K. Also, all video sequences were encoded using four frame rates up to 120 fps, presenting different temporal resolutions (TR). The Structural Similarity Index Measure (SSIM) is used as an objective video quality metric. After data augmentation, we propose NN models that estimate optimal TR and bitrate values as switching points to a higher SR. These results can be further used as input parameters for the bitrate ladder construction for video sequences of a certain complexity.

A mixed-method exploration of #vapingcessation videos on TikTok.

E-cigarette use has risen rapidly, especially among young people. TikTok is one of the most commonly used social media platforms among young people. The representation of vaping cessation messages in TikTok content remains unexplored. We aimed to address this gap by analyzing the vaping cessation content prevalent on TikTok. This cross-sectional mixed methods study analyzed the top 100 TikTok videos-determined based on view-counts-related to vaping cessation. Our quantitative analysis examined video characteristics, while our qualitative analysis identified key thematic representations. The final sample included 87 videos, with a combined total of over 100 billion views. Most videos were created by regular users (general TikTok content creators) rather than those identifying as health experts or professionals such as doctors, acupuncturists or dental hygienists. Very few posts mentioned evidence-based cessation aids or support. Videos scored high on understandability but low on actionability based on standardized criteria. Qualitative analysis identified nine key themes: motivations for quitting, barriers to quitting, use of aids and strategies, relapse experiences, identity and self-image, celebrating cessation success, emotional journey of quitting vaping, seeking social support, and limited mentions of healthcare providers. TikTok appears to be a useful influential forum for studying quitting behavior through the sharing of personal experiences and perspectives on vaping cessation. Opportunities exist to integrate evidence-based messaging and support into this platform. Findings can inform social media interventions designed to counter pro-vaping content and promote quitting e-cigarette use. Partnerships among health professionals, public health experts, and social media influencers may be leveraged to disseminate proven vaping cessation techniques on leading social media platforms such as TikTok.

Enhanced Infant Movement Analysis Using Transformer-Based Fusion of Diverse Video Features for Neurodevelopmental Monitoring

Neurodevelopment is a highly intricate process, and early detection of abnormalities is critical for optimizing outcomes through timely intervention. Accurate and cost-effective diagnostic methods for neurological disorders, particularly in infants, remain a significant challenge due to the heterogeneity of data and the variability in neurodevelopmental conditions. This study recruited twelve parent–infant pairs, with infants aged 3 to 12 months. Approximately 25 min of 2D video footage was captured, documenting natural play interactions between the infants and toys. We developed a novel, open-source method to classify and analyse infant movement patterns using deep learning techniques, specifically employing a transformer-based fusion model that integrates multiple video features within a unified deep neural network. This approach significantly outperforms traditional methods reliant on individual video features, achieving an accuracy of over 90%. Furthermore, a sensitivity analysis revealed that the pose estimation contributed far less to the model’s output than the pre-trained transformer and convolutional neural network (CNN) components, providing key insights into the relative importance of different feature sets. By providing a more robust, accurate and low-cost analysis of movement patterns, our work aims to enhance the early detection and potential prediction of neurodevelopmental delays, whilst providing insight into the functioning of the transformer-based fusion models of diverse video features.

Evaluating the quality and reliability of oncologic breast reconstruction videos on youtube

Oncologic breast reconstruction (OBR) is a complex process that requires consideration of multiple factors, including chemoradiation, extent of cancer treatment, and surgical approach. Patients often feel uncertain about the numerous surgical options and may turn to popular social media platforms like YouTube for information. Thus, this study aims to assess the quality and reliability of YouTube videos related to OBR.We conducted a retrospective, cross-sectional analysis of YouTube videos related to OBR. Search terms were obtained from plasticsurgery.org and Google Trends. The first ten videos for each search term were analyzed. Videos were categorized by source and subject matter and independently reviewed by three evaluators using the DISCERN scale.This study examined 172 YouTube videos. Five video source categories were identified: Health Care Administrations, Physicians, Non-physician Providers, News Organizations, and Patients. Health Care Administration accounts received the highest overall DISCERN score of 3.6 ± 0.51, followed by Physicians at 2.98 ± 0.93, with News Organization accounts scoring the lowest at 2.22 ± 0.60 (p < 0.001). Videos from academic sources (Physicians and Health Care Administrators) had higher DISCERN scores compared to non-academic sources (News Organizations and Patients), 2.98 ± 0.95 versus 2.28 ± 0.77, respectively (p < 0.001).Our findings indicate that videos from academic sources generally exhibit higher DISCERN scores, pointing to a higher content quality and reliability standard. Given the increasing reliance on YouTube for healthcare information, our study underscores the need for healthcare professionals to engage more actively in content creation and dissemination.

Query-aware multi-scale proposal network for weakly supervised temporal sentence grounding in videos

Recently, weakly supervised temporal sentence grounding in videos (TSGV) has attracted extensive attention because it does not require precise start-end time annotations during training, and it can quickly retrieve interesting segments according to user needs. In weakly supervised TSGV, query reconstruction (QR)-based methods are the current mainstream, and the quality of proposals determines their performance. QR-based methods have two problems in proposal quality. First, a multi-modal global token is usually mapped to proposals with limited duration diversity, making it difficult to capture relevant segments at varying durations in real scenarios. Additionally, Gaussian functions are typically used to generate relatively fixed weights for frames within proposals, which weigh the original video features to generate proposal-specific features. This results in query-irrelevant frames affecting the discrimination of the proposal features. In this study, we propose a query-aware multi-scale proposal network (QMN). Initially, pre-trained encoders are used to extract video and query features. Subsequently, a multi-scale proposal generation module is designed to refine video features guided by queries and diversify the duration of the proposal. This module performs multi-modal interaction and multi-scale modeling to obtain proposals of different durations. Furthermore, to extract discriminative proposal features and enhance the modeling of proposal frame correlation, a query-aware weight generator is constructed to learn frame weights to suppress query-irrelevant frame representations through contrastive learning. Finally, the masked query is reconstructed using the proposal features to select the best proposal. The effectiveness of the proposed QMN is verified through experiments on the Charades-STA and ActivityNet-Captions datasets.

Medical misinformation in social media: Representations of gastrointestinal disorders on a short video platform

Background and Objectives: Short video platforms have become one of the most common methods for disseminating medical information on social media. We analysed gastrointestinal (GI)-related content on TikTok, focusing on the creators’ background, patterns of content utilisation and overall content quality and understandability, using validated metrics to quantify the degree of misinformation. Methods: Twelve hashtags related to common GI conditions on TikTok and the 20 most engaging posts associated with each hashtag were selected for review. The ‘most engaging’ status was determined by the application’s internal algorithm. Six GI board-certified physicians evaluated the videos using two validated instruments: DISCERN quality criteria and Patient Education Materials Assessment Tool (PEMAT-A/V) survey. Results: In total, 253 videos were reviewed with 18.6 billion likes, 191,000 comments, 935,000 saves and 927,000 shares. IBS hashtag had the greatest number of views (2.2 billion), while videos on haemorrhoids had the most likes (4.3 million), comments (59,000) and shares (288,000). Most videos 177/253 (70%) were created by non-healthcare workers. The median video quality was 2/5 on DISCERN Likert-type scale with 23/75 DISCERN score. Of 186 videos, 114 (61.3%) were rated ‘very poor’; 58/186 (31.2%) ‘poor’, 13/186 (7%) ‘fair’ and only one video (0.5%) was rated ‘excellent’ on DISCERN grading system. The median PEMAT-A/V scores were 7/12 (73.9%) for understandability and 1/3 (33.3%) for actionability. Conclusion: Our study highlights significant public interest in common GI disorders. However, most posts lacked substantive information and did not have verifiable sources.

Towards fine-grained adaptive video captioning via Quality-Aware Recurrent Feedback Network

To achieve accurate video captioning, most existing solutions focus on employing auxiliary features to provide complementary generation clues, such as detected target regions, audio speech signals, and in-plane text. However, beyond introducing external information, the potential of using the generated captions has been overlooked. We argue that the intrinsic value lies in the caption text, which can faithfully reflect the video content. To this end, we propose to reuse the captions via a Quality-Aware Recurrent Feedback Network (QARFNet), a model that progressively exploits the supportive power between video features and the generated captions. Specifically, inspired by the backtracking AdaBoost algorithm, we propose to build a recurrent loop structure that recycles the obtained linguistic predictions to refine the input visual features. Referring to the greedy selective nature of the AdaBoost algorithm, we set a quality-aware gate to assess the necessity of forwarding the next loop, lessening the complexity. After the refinement loop, if the confidence score remains below the predefined threshold, we select the output with the highest confidence score to advance the subsequent transformer decoder. To facilitate feedback in each loop, a Multi-level Update Module is constructed before fusing linguistic predictions with video features. This involves extracting nouns and verbs from linguistic predictions to lighten relevant video tokens, achieving self-interactions across recurrent loops. By merging the coarse-grained and gradually refined linguistic prediction features, the video tokens are semantically closer to the desired textual representation. To alleviate the increased calculation caused by the loops, we employ a flat transformer encoder to decrease the complexity. The experimental results on several benchmark datasets confirm the effectiveness of our approach, with superior performance compared to the state-of-the-art.

Dynamic assessment of visual fatigue during video watching: Validation of dynamic rating based on post-task ratings and video features

People watching video displays for long durations experience visual fatigue and other symptoms associated with visual discomfort. Fatigue-reduction techniques are often applied but may potentially degrade the immersive experience. To appropriately adjust fatigue-reduction techniques, the changes in visual fatigue over time should be analyzed which is crucial for the appropriate adjustment of fatigue-reduction techniques. However, conventional methods used for assessing visual fatigue are inadequate because they rely entirely on post-task surveys, which cannot easily determine dynamic changes. This study employed a dynamic assessment method for evaluating visual fatigue in real-time. Using a joystick, participants continuously evaluated subjective fatigue whenever they perceived changes. A Simulator Sickness Questionnaire (SSQ) validated the results, which indicated significant correlations between dynamic assessments and the SSQ across five items associated with symptoms associated with visual discomfort. Furthermore, we explored the potential relationship between dynamic visual fatigue and objective video features, e.g., optical flow and the V-values of the hue/saturation value (HSV) color space, which represent the motion and brightness of the video. The results revealed that dynamic visual fatigue significantly correlated with both the optical flow and the V-value. Moreover, based on machine learning models, we determined that the changes in visual fatigue can be predicted based on the optical flow and V-value. Overall, the results validate that dynamic assessment methods can form a reliable baseline for real-time prediction of visual fatigue.

Unlocking TikTok’s Formula: Video Length, Tags, and Beauty Content Success

With the rapid growth of TikTok in the digital era, beauty influencers have gained significant influence over consumer decisions, particularly through makeup product reviews. TikTok’s short and interactive content style has brought attention to how video characteristics such as length and the number of tags affect engagement metrics like views, likes, and comments. This research explores the impact of these factors using data analysis from a sample of 25 TikTok videos, where the average video length was 2 minutes and 20 seconds, and the top 25% of videos had lengths exceeding 3 minutes and 30 seconds. The findings reveal that longer videos tend to foster deeper engagement by providing more comprehensive reviews, with the top 25% of videos achieving 30% higher engagement than the average. Additionally, a higher number of tags significantly increases video reach. These insights are valuable for influencers and marketers looking to optimize their content strategies, particularly in the beauty industry, where maximizing audience engagement is crucial.

S583 Viewer Engagement and Content Quality in Eosinophilic Esophagitis Videos: Evaluating Academic and Private Contributions on YouTube

S583 Viewer Engagement and Content Quality in Eosinophilic Esophagitis Videos: Evaluating Academic and Private Contributions on YouTube

Assessment of reliability and information quality of youtube videos about dietary recommendations for irritable bowel syndrome

Assessment of reliability and information quality of youtube videos about dietary recommendations for irritable bowel syndrome

YouTube videos on gingival grafting procedures: Content and quality analysis

BackgroundPeriodontal surgery for gingival defects is widely recognized by dental care professionals and researchers for its effectiveness in treating gingival recession and improving oral health outcomes. YouTube (Google LLC) is 1 of the health information sources patients and clinicians use, and assessing its content quality is crucial. The authors aimed to examine the content and quality of YouTube videos on gingival graft procedures. MethodsThe online video streaming platform YouTube was searched using the key word gingival graft. Two independent examiners analyzed a total of 120 videos; a third examiner assessed interrater reliability. Fifty videos that met the inclusion criteria were included in the study. The assessed content topics for these YouTube videos consisted of 13 different categories, and their overall quality was evaluated using the Video Information and Quality Index (VIQI). Statistical analyses were performed using SAS software, Version 9.4 (SAS Institute). ResultsThere were 23 videos in the high-quality content groups and 27 videos in the low-quality content group. Hospitals and universities uploaded most of the included videos. The most commonly discussed topics in the included videos were the patient’s condition (36 [72%]) and area of tissue graft (34 [68%]). The total VIQI score and flow had a significant impact on the overall content score (P < .05). ConclusionsThere was a direct correlation between total VIQI scores and total content scores and an inverse relationship between viewing rate and total content scores. Practical ImplicationsTo ensure patients receive accurate and up-to-date information about treatment, the authors recommend guiding them toward reliable resources by means of providing direct links to trustworthy websites, creating and sharing playlists of reliable educational videos, and offering printed materials with quick-response codes linking to verified sources. These actions will help patients easily access and trust the information they need for their treatment decisions.

S899 The Role of Video Quality in Central Reader Disagreement in Assessment of the Endoscopic Mayo Score in Ulcerative Colitis Trials

S899 The Role of Video Quality in Central Reader Disagreement in Assessment of the Endoscopic Mayo Score in Ulcerative Colitis Trials

Development and dissemination of a series of surgical skills and procedures video tutorials using a novel, low-cost, and sustainable simulation kit (GlobalSurgBox)

IntroductionSurgical simulation and video-based learning are limited in lower-resource settings. We sought to develop and assess a series of surgical tutorials using a low-cost simulator. MethodsWe created 8 surgical skills and procedures videos using low-cost equipment. We assessed video quality using the DISCERN scale and the Global Quality Scale (GQS). ResultsVideos ranged from surgical techniques to complex procedures. We uploaded these to Youtube and included them in the curriculum of a medical school in Rwanda. Excluding the cost of the kit (25 USD), production costs ranged from 2 to 5 USD. All videos scored a mean DISCERN of 2.44 ​± ​1.05 and GQS of 3.06 ​± ​0.90. Generally, these lacked points on providing additional sources of information and addressing areas of uncertainty. ConclusionsThis study addresses the demand for accessible surgical education resources. Using low-cost, standardized materials ensures consistency, democratization of training, and feasibility.

FReD-ViQ: Fuzzy Reinforcement Learning Driven Adaptive Streaming Solution for Improved Video Quality of Experience

FReD-ViQ: Fuzzy Reinforcement Learning Driven Adaptive Streaming Solution for Improved Video Quality of Experience

Qualitative analysis of Portuguese-language YouTube videos about autism spectrum disorders

BackgroundIn the digital age, social media platforms such as YouTube have become significant channels for disseminating health information, including content related to autism spectrum disorder (ASD). The quality and reliability of this information, especially when produced by healthcare professionals, are crucial for public health education and promotion. This study aims the content of Portuguese-language videos about the treatment of ASD on YouTube, produced by healthcare providers from 2019 to 2023, assessing their quality and alignment with evidence-based practices. MethodsA qualitative exploratory descriptive approach was used, with content analysis based on Bardin's method. A total of 41 videos were selected using keywords related to ASD. Transcriptions were analyzed for discussions on treatment approaches, best practices, and professional recommendations according to DSM-V and ICD-10 guidelines. The quality of information was assessed using the DISCERN questionnaire. ResultsThe analysis revealed significant variability in the quality of the information. Videos were categorized into four quality groups based on DISCERN scores: good (n = 6), moderate (n = 11), poor (n = 20), and very poor (n = 4). Good quality videos had the highest engagement metrics and overall quality scores. Common themes identified included defining and understanding ASD, ABA interventions and strategies, family and social impact, skills development, and challenges and solutions. ConclusionWhile some videos provided accurate, evidence-based information, a substantial portion did not meet minimum quality criteria. This highlights the need for improved mechanisms to ensure the dissemination of reliable health information on social media platforms.

STMixer: A One-Stage Sparse Action Detector.

Traditional video action detectors typically adopt the two-stage pipeline, where a person detector is first employed to generate actor boxes and then 3D RoIAlign is used to extract actor-specific features for action recognition. This detection paradigm requires multi-stage training and inference, and the feature sampling is only constrained inside the box, failing to effectively leverage richer context information outside. Recently, several query-based action detectors have been proposed to predict action instances in an end-to-end manner. However, they still lack adaptability in feature sampling and decoding, thus suffering from the issues of inferior performance or slower convergence. In this paper, we propose two core designs for a more flexible one-stage sparse action detector. First, we present a query-based adaptive feature sampling module, which endows the detector with the flexibility of mining a group of discriminative features from the entire spatio-temporal domain. Second, we devise a decoupled feature mixing module, which dynamically attends to and mixes video features along the spatial and temporal dimensions respectively for better feature decoding. Based on these designs, we instantiate two detection pipelines, that is, STMixer-K for keyframe action detection and STMixer-T for action tubelet detection. Without bells and whistles, our STMixer detectors obtain the state-of-the-art results on five challenging spatio-temporal action detection benchmarks for keyframe action detection or action tube detection.