Video Modeling Research Articles

Implementation of Video Modeling Animation on Toilet Training Ability in Autism Spectrum Disorders

INTRODUCTION: Toilet training is critical for the quality of life of individuals with autism spectrum disorders (ASD) and other developmental disabilities. This is essential because an individual with ASD is often late to acquire these skills or fails to do so. This study aims to identify and analyze the implementation of animation video modeling (AVM) to improve toilet training abilities in children with autism spectrum disorders. MATERIALS AND METHODS: The research uses a quantitative descriptive approach with an experimental research design using One Group Pretest Posttest Design. The pretest was carried out before the intervention, and the posttest was carried out after the AVM gave toilet training intervention. The sampling technique was in the form of total sampling, namely 50 children(43 boys and 7 girls) with autism. The research instrument was a checklist of statements. Data analysis using the Wilcoxon Sign Rank Test. RESULTS: The results showed that almost half of the ability levels of children with autism before being given AVM toilet training were in a suitable category, nearly half of the children with autism after being given treatment were in a good type with a p-value 0.011; there was a significant difference in the level of ability before and after given an AVM. CONCLUSION: AVM is one of the effective interventions to improve toilet training abilities in children with ASD. AVM can help children in the elimination process that can be done independently in improving the quality of life in children with autism.

Effects of a video modeling intervention to improve hand washing independence in children with disabilities

Effects of a video modeling intervention to improve hand washing independence in children with disabilities

비디오 모델링과 연합한 자기점검 전략이 지적장애 초등학생의 식사시간 활동에 미치는 영향

The purpose of this study was to examine video modeling combined with self-monitoring strategies on meal time activities which are meal preparation, cleaning the table, tooth brushing skills of an elementary school student with intellectual disabilities in a special school. The research questions were as follows: First, does the video modeling combined with self-monitoring strategies affect on meal time activities of an elementary school student with intellectual disabilities? Secondly, does video modeling combined with self-monitoring affect its maintenance and generalization of meal time activities for an elementary school student with intellectual disabilities? One elementary school student with intellectual disabilities, age 11, participated in the study. Multiple probe design across behaviors were utilized for this study. The result of the study were as follows: video modeling combined with self-monitoring improved the meal time activities of the participant. The effects were maintained three weeks after the intervention. The average of non-overlapping data point(PND) was 96%, indicating a ‘high effective level’ of intervention effects. The result of this study proved that video modeling combined with self-monitoring strategies was an effective intervention enhancing meal time activities for an elementary school student with intellectual disabilities. The results were maintained and generalized at home.

Knowledge Distillation in Video-Based Human Action Recognition: An Intuitive Approach to Efficient and Flexible Model Training.

Training a model to recognize human actions in videos is computationally intensive. While modern strategies employ transfer learning methods to make the process more efficient, they still face challenges regarding flexibility and efficiency. Existing solutions are limited in functionality and rely heavily on pretrained architectures, which can restrict their applicability to diverse scenarios. Our work explores knowledge distillation (KD) for enhancing the training of self-supervised video models in three aspects: improving classification accuracy, accelerating model convergence, and increasing model flexibility under regular and limited-data scenarios. We tested our method on the UCF101 dataset using differently balanced proportions: 100%, 50%, 25%, and 2%. We found that using knowledge distillation to guide the model's training outperforms traditional training without affecting the classification accuracy and while reducing the convergence rate of model training in standard settings and a data-scarce environment. Additionally, knowledge distillation enables cross-architecture flexibility, allowing model customization for various applications: from resource-limited to high-performance scenarios.

Follow Your Pose: Pose-Guided Text-to-Video Generation Using Pose-Free Videos

Generating text-editable and pose-controllable character videos have an imperious demand in creating various digital human. Nevertheless, this task has been restricted by the absence of a comprehensive dataset featuring paired video-pose captions and the generative prior models for videos. In this work, we design a novel two-stage training scheme that can utilize easily obtained datasets (i.e., image pose pair and pose-free video) and the pre-trained text-to-image (T2I) model to obtain the pose-controllable character videos. Specifically, in the first stage, only the keypoint image pairs are used only for a controllable text-to-image generation. We learn a zero-initialized convolutional encoder to encode the pose information. In the second stage, we finetune the motion of the above network via a pose-free video dataset by adding the learnable temporal self-attention and reformed cross-frame self-attention blocks. Powered by our new designs, our method successfully generates continuously pose-controllable character videos while keeps the editing and concept composition ability of the pre-trained T2I model. The code and models are available on https://follow-your-pose.github.io/.

ViSTec: Video Modeling for Sports Technique Recognition and Tactical Analysis

The immense popularity of racket sports has fueled substantial demand in tactical analysis with broadcast videos. However, existing manual methods require laborious annotation, and recent attempts leveraging video perception models are limited to low-level annotations like ball trajectories, overlooking tactics that necessitate an understanding of stroke techniques. State-of-the-art action segmentation models also struggle with technique recognition due to frequent occlusions and motion-induced blurring in racket sports videos. To address these challenges, We propose ViSTec, a Video-based Sports Technique recognition model inspired by human cognition that synergizes sparse visual data with rich contextual insights. Our approach integrates a graph to explicitly model strategic knowledge in stroke sequences and enhance technique recognition with contextual inductive bias. A two-stage action perception model is jointly trained to align with the contextual knowledge in the graph. Experiments demonstrate that our method outperforms existing models by a significant margin. Case studies with experts from the Chinese national table tennis team validate our model's capacity to automate analysis for technical actions and tactical strategies. More details are available at: https://ViSTec2024.github.io/.

New Metrics and Dataset for Biological Development Video Generation

Image generative models have advanced in many areas to produce synthetic images of high resolution and detail. This success has enabled its use in the biomedical field, paving the way for the generation of videos showing the biological evolution of its content. Despite the power of generative video models, their use has not yet extended to time-based development, focusing almost exclusively on generating motion in space. This situation is largely due to the lack of specific data sets and metrics to measure the individual quality of videos, particularly when there is no ground truth available for comparison. We propose a new dataset, called GoldenDOT, which tracks the evolution of apples cut in parallel over 10 days, allowing to observe their progress over time while remaining static. In addition, four new metrics are proposed that provide different analyses of the generated videos as a whole and individually. In this paper, the proposed dataset and measures are used to study three state of the art video generative models and their feasibility for video generation with biological development: TemporalGAN (TGANv2), Low Dimensional Video Discriminator GAN (LDVDGAN), and Video Diffusion Model (VDM). Among them, the TGANv2 model has managed to obtain the best results in the vast majority of metrics, including those already known in the state of the art, demonstrating the viability of the new proposed metrics and their congruence with these standard measures.

The effectiveness of video model with Bug-in-Ear coaching in the ınstruction of employment skills to adolescents with autism spectrum disorder

Objectives Employment skills enable individuals to progress in social, psychological, economic, and many other ways. Employment skills are among the skills that enable individuals with autism to live independently and improve their quality of life. However, despite this importance, it is one of the skills that individuals with autism have problems with. Methods In the study, the effectiveness of the instruction of employment skills with the video model (VM) presented with Bug in Ear (BIE) coaching was tested with the repeated multiple probe model and inter-skill probe trial, a single-subject research model. Results The findings demonstrated that the participants acquired the employment skills to 100% proficiency/mastery, retained these skills to 100% proficiency after 1, 5, and 12 wk, and could generalize these skills to different individuals, environments, and tools. Pre-intervention data showed an increase compared to the post-intervention. Social validity findings for the young adults with autism spectrum disorder (ASD) and their parents were positive. Conclusions In this research, it has been seen that the VM with BIE is effective in teaching young adults with ASD the skills of ‘preparing a CV, applying for a job to Turkish Employment Agency (İŞKUR) and participating in a job interview’ from employment skills. Although it may seem difficult to teach individuals with autism, it seems that they can learn employment-related skills with evidence-based practices.

Backdoor Two-Stream Video Models on Federated Learning

Video models on federated learning (FL) enable continual learning of the involved models for video tasks on end-user devices while protecting the privacy of end-user data. As a result, the security issues on FL, e.g., the backdoor attacks on FL and their defense have increasingly becoming the domains of extensive research in recent years. The backdoor attacks on FL are a class of poisoning attacks, in which an attacker, as one of the training participants, submits poisoned parameters and thus injects the backdoor into the global model after aggregation. Existing backdoor attacks against videos based on FL only poison RGB frames, which makes that the attack could be easily mitigated by two-stream model neutralization. Therefore, it is a big challenge to manipulate the most advanced two-stream video model with a high success rate by poisoning only a small proportion of training data in the framework of FL. In this paper, a new backdoor attack scheme incorporating the rich spatial and temporal structures of video data is proposed, which injects the backdoor triggers into both the optical flow and RGB frames of video data through multiple rounds of model aggregations. In addition, the adversarial attack is utilized on the RGB frames to further boost the robustness of the attacks. Extensive experiments on real-world datasets verify that our methods outperform the state-of-the-art backdoor attacks and show better performance in terms of stealthiness and persistence.

KNOWLEDGE OF USING DIGITAL TECHNOLOGIES IN PHYSICAL EDUCATION OF STUDENTS

Relevance: today, digital technologies have penetrated into all spheres of human activity, including education. Scientists and experts point out that the potential of digital technologies in physical education of students is not fully realized. Objective: to search and analyze scientifi c knowledge about the knowledge of using digital technologies in physical education of students. Materials and methods: structured search for current information in scientifi c databases (WoS, Scopus, PubMed, eLIBRARY) and survey of a focus group composed of university teachers and students. Results and conclusion: it was founded that technologies associated with digital learning resources (online courses) and media content (social networks and mobile applications) are most often used in physical education classes. At the same time, students want to make more active use of technologies related to both digital video modeling and gamifi ca-tion (exergames).

Preschool Special Educators’ Exposure to and Use of Evidence-Based and Low-Value Practices for Children with Autism

Exploring what practices early childhood special education (ECSE) teachers have been exposed to and use may highlight the need for improved dissemination and implementation supports. In this survey study, 312 U.S. ECSE teachers reported exposure (familiarity, training) and use of evidence-based practices (EBPs) and low-value practices (LVPs) for children with autism. Teachers reported high familiarity, training, and use of many EBPs (e.g., modeling, reinforcement), while other EBPs remain minimally used despite high familiarity (video modeling; 23.4%). Some teachers reported not using EBPs they had been trained in. Overall, teachers reported some familiarity and use of LVPs ( m = 20.47, 0%–50.64%) and minimal training (preservice = 1.6%–33.33%). Research should explore exposure and other possible explanations of the limited use of effective practices.

Eating skills in children with autism via video modeling: an evaluation of face and content validation

<p>Children with autism spectrum disorder (ASD) may have their eating habits impacted by their traits, such as difficulty remaining seated at the table and exhibiting behaviors during mealtime. This often worries parents and causes stress and anxiety. Given video modeling's potential in helping children with ASD, experts should evaluate its effectiveness for teaching eating skills due to limited research in this area. Thus, the study aim to evaluate the validity of the developed videos of eating skills for children with ASD. For face validity, a focus group discussion was held to obtain experts' feedback on the eating skills video and was transcribed verbatim. Otherwise for content validity, the experts received content validity assessment forms and guidelines via email and the results was analysed using Microsoft excell. Face validity revealed four themes: suitability of videos, video quality and clarity, video instructions suitability, task analysis arrangement, and "other" recommendations. On the other hand, Video 1 and Video 2 scored 0.93 for relevance, clarity, simplicity, and ambiguity for the S-CVI/Ave (scale-level content validity index based on the average method) which was considered acceptable. Video 3 received a satisfactory S-CVI/Ave score of 0.93 for relevance, simplicity, and ambiguity, but its clarity score at 0.91, and still within an acceptable range, though not particularly high. As a conclusion, all task analyses for the videos met the requirement to be considered acceptable. Addressing the suggestions for improvement can increase the video's efficacy in teaching eating skills and relieving parents' concerns.</p>

Video quality adaptation using CNN and RNN models for cost-effective and scalable video streaming Services

Video streaming services require adaptive bit rate strategies that optimize video quality based on network conditions and user preferences to provide a cost-effective and scalable solution. In this manuscript, we present a novel architecture that utilizes a combination of Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) to extract features from the video stream and predict optimal bit rates for each frame. The CNN feature extractor extracts relevant features from the input video stream, which are then passed on to the RNN predictor, where the optimal bit rate is predicted to ensure a personalized viewing experience tailored to the current network conditions and specific user preferences. Our experimental results demonstrate that the proposed architecture achieves significant improvements in video quality while minimizing bandwidth usage and providing a better user experience. Specifically, our results show a 37.1% improvement in average bit rate, indicating that the proposed architecture optimizes the video quality and reduces bandwidth usage by 37.1% on average. We also observed a 16.6% improvement in Quality of Experience (QoE), meaning that users will perceive the video quality as better. Additionally, rebuffering was reduced by 87.5%, indicating that users enjoy smoother video playback without interruptions. Importantly, our architecture is optimized for Dynamic Adaptive Streaming over HTTP (DASH), addressing the need for efficient streaming over the most widely used protocol in the industry.

Effects of Self and Peer Video Modeling on Textual Responses in Children with ASD

A videomodelação consiste na exibição de vídeos retratando um comportamento para instalar ou modificar o respectivo comportamento do espectador, que, após assistir ao vídeo, é solicitado a imitá-lo. O presente estudo comparou videomodelação por si mesmo e videomodelação por par para o ensino de respostas textuais sob controle de consoantes do alfabeto português em três crianças diagnosticadas com Transtorno do Espectro Autista (TEA). Para isso, foi utilizado um delineamento misto de linha de base múltipla entre participantes e de tratamentos alternados. Alguns cuidados no controle experimental foram acrescentados em relação a trabalhos anteriores: as consequências para a emissão de respostas corretas e incorretas nas condições de linha de base e de tratamento foram idênticas; os pares utilizados foram fisicamente semelhantes aos participantes; e incluiu-se a avaliação de preferência dos participantes pelos vídeos. Os resultados indicaram efeitos semelhantes do ensino em ambas as condições de videomodelação, sendo que dois dos três participantes aprenderam igualmente nas duas condições e um participante atingiu o critério apenas na videomodelação por par. As evidências sugerem efeitos positivos advindos do tipo de intervenção avaliado e a necessidade de futuras pesquisas avaliarem o ensino de outras habilidades.

An evaluation of training components necessary to teach staff members to conduct an intensive pediatric feeding intervention

AbstractIn applied behavior analysis, parents or other caregivers are often trained to implement an intervention after the intervention has been demonstrated effective when implemented by professionals. However, training novice staff members to conduct specific behavioral interventions has received less attention. This is particularly true in the context of training staff to implement interventions to treat food refusal. In the current study, we implemented three levels of training to evaluate the level of intensity required to train staff members (i.e., behavior technicians, Registered Behavior Technicians®, and Board‐Certified Behavior Analysts®) to implement a feeding protocol with a role play partner. Only one participant met the mastery criteria following the first training phase, which included written instructions and video modeling. Following 72 h of exposure to the instructional materials, we provided participants an opportunity to ask clarifying questions about the procedure (second training phase). Only one participant met mastery in this phase. Three other participants required in vivo feedback and modeling (third training phase) to master the protocol. We discuss clinical implications and future research directions.

Exploring the Acquisition of Social Communication Skills in Children with Autism: Preliminary Findings from Applied Behavior Analysis (ABA), Parent Training, and Video Modeling.

Social communication skills, especially eye contact and joint attention, are frequently impaired in autism spectrum disorder (ASD) and predict functional outcomes. Applied behavior analysis is one of the most common evidence-based treatments for ASD, but it is not accessible to most families in low- and middle-income countries (LMICs) as it is an expensive and intensive treatment and needs to be delivered by highly specialized professionals. Parental training has emerged as an effective alternative. This is an exploratory study to assess a parental intervention group via video modeling to acquire eye contact and joint attention. Four graded measures of eye contact and joint attention (full physical prompt, partial physical prompt, gestural prompt, and independent) were assessed in 34 children with ASD and intellectual disability (ID). There was a progressive reduction in the level of prompting required over time to acquire eye contact and joint attention, as well as a positive correlation between the time of exposure to the intervention and the acquisition of abilities. This kind of parent training using video modeling to teach eye contact and joint attention skills to children with ASD and ID is a low-cost intervention that can be applied in low-resource settings.

Feasibility Test of Web-Based Video Modelling Media for Learning Social Skills for Autistic Students at Inclusive Elementary Schools in Surakarta

<p><em>This research aims to know the feasibility level of Web-based Video Modelling Media for Social Skills of Autistic Students in Inclusive Elementary School in Surakarta. The current study used quantitative design as the research method. The test procedure used is Alpha Testing and Beta Testing. The subject of this research were autistic students at Inclusive Elementary School in Surakarta. The technique of data collection is done by a questionnaire. Data analysis using quantitative data analysis. The research results obtained were that Web-based Video Modeling Media for Social Skills of Autistic students received an assessment of 83.7% from media experts, 87.47% from material experts, and 88.5% from from IT experts and 89.01% from users candidate. So it can be infered that video modeling media for the social skills of autistic students is very decent for used to learning the social skills of autistic students in inclusive elementary schools in Surakarta</em></p>

Video Modeling and Scenario-Based Learning on Interpersonal Skills and Work Readiness for Transition Age Students with Disabilities: An Interdisciplinary Approach

ABSTRACT This pretest/posttest study investigated the effectiveness of an interdisciplinary program in which occupational therapists and speech-language pathologists used a combination of video modeling, scenario-based learning, guided discussion, and reflection on the development of interpersonal skills and work readiness among students with disabilities. This study examined 20 high school and postgraduate transitional students enrolled full-time in Life Skills Support classrooms in two rural high schools. Participants completed 21 dynamic and interactive group sessions which included video modeling, scenario-based learning, guided discussion, and reflection. Sessions were completed once a week, for one hour, and focused on enhancing interpersonal skills, cognitive skills, work habits, and performance skills. Sessions were delivered via an interdisciplinary model with participants’ primary occupational therapist and speech-language pathologist. Results of the paired-samples t-test found a statistically significant increase in the Becker Work Adjustment Profile-2 scores in three categories: Cognitive Skills, Work Performance, and Broad Work Adjustment. All students showed growth on Goal Attainment Scaling (GAS) T-scores; 18 of the 20 students achieved GAS T-scores greater than threshold performance.

Comparing teaching with tactile guidance with video modeling with self-evaluative video feedback to train medical skills.

Modern medical training consists largely of lecture-based instruction and in vivo or video modeling of specific skills. Other instructional methods, such as teaching with acoustical guidance (TAGteach), have rarely been evaluated. In this study, we compared teaching with tactile guidance, or tactile TAGteach in which a vibratory stimulus is delivered to indicate a correct response, with video modeling and self-evaluative video feedback to teach four participants two medical skills: simple interrupted suture and endotracheal intubation. The results showed that both instructional methods improved performance. However, three participants met the mastery criterion in the tactile TAGteach condition first, although this instructional method required more time to train the skills. We discuss the implications of these findings for training skills to medical practitioners.

An Effective Video Transformer With Synchronized Spatiotemporal and Spatial Self-Attention for Action Recognition.

Convolutional neural networks (CNNs) have come to dominate vision-based deep neural network structures in both image and video models over the past decade. However, convolution-free vision Transformers (ViTs) have recently outperformed CNN-based models in image recognition. Despite this progress, building and designing video Transformers have not yet obtained the same attention in research as image-based Transformers. While there have been attempts to build video Transformers by adapting image-based Transformers for video understanding, these Transformers still lack efficiency due to the large gap between CNN-based models and Transformers regarding the number of parameters and the training settings. In this work, we propose three techniques to improve video understanding with video Transformers. First, to derive better spatiotemporal feature representation, we propose a new spatiotemporal attention scheme, termed synchronized spatiotemporal and spatial attention (SSTSA), which derives the spatiotemporal features with temporal and spatial multiheaded self-attention (MSA) modules. It also preserves the best spatial attention by another spatial self-attention module in parallel, thereby resulting in an effective Transformer encoder. Second, a motion spotlighting module is proposed to embed the short-term motion of the consecutive input frames to the regular RGB input, which is then processed with a single-stream video Transformer. Third, a simple intraclass frame interlacing method of the input clips is proposed that serves as an effective video augmentation method. Finally, our proposed techniques have been evaluated and validated with a set of extensive experiments in this study. Our video Transformer outperforms its previous counterparts on two well-known datasets, Kinetics400 and Something-Something-v2.