异构复合迁移学习的视频内容标注方法

针对传统的机器学习需要大量的人工标注训练模型的弊端,以及目前多数迁移学习方法只适用于同构空间的问题,提出了一种异构复合迁移学习（HCTL）的视频内容标注方法。首先,借助视频与图像的对应关系,利用典型相关性分析（CCA）来实现图像域（源域）和视频域（目标域）特征空间的同构化;然后,基于这两个特征空间向共同空间投影的代价最小化这一思想,找到源域特征空间向目标域特征空间对齐的矩阵;最后,通过对齐矩阵使得源域特征能够翻译到目标域特征空间中去,进而实现知识迁移,完成视频内容标注任务。所提方法在Kodak数据库上的平均标注准确率达到了35.81%,与标准的支持向量机（S-SVM）领域适应支持向量机（DASVM）、异构直推式迁移学习（HTTL）、跨领域的结构化模型（CDSM）、领域选择机（DSM）、异构源域下的多领域适应（MDA-HS）和判别性相关分析（DCA）方法相比分别提高了58.03%、23.06%、45.04%、6.70%、15.52%、13.07%和6.74%;而在哥伦比亚用户视频（CCV）数据库上达到了20.73%,分别相对提高了133.71%、37.28%、14.34%、24.88%、16.40%、20.73%和12.48%。实验结果表明先同构再对齐的复合迁移思想在异构领域适应问题上能够有效地提升识别准确率。     

Examining different types of collaborative learning in a complex computer-based environment: A cognitive load approach

This study compared the effects of two collaborative learning strategies (Open-ended and Task-based) with an individualized learning strategy on individual learning in a computer-based environment. The experiment sought ecological validity by conducting it under real teaching and homework conditions. Ninety-four students from grade 9 participated in a webpage design task. Cognitive load theory was used to predict that the collaborative approaches would outperform the individualized approach due to reduced cognitive load. This hypothesis was confirmed by performance scores and cognitive load only in the case of the Open-ended collaborative learning condition. Evidence was also found that the Open-ended collaborative learning condition outperformed the Task-based collaborative one. It was concluded that in collaborative learning a more Open-ended task design together with moderate independent sub-task requirements leads to more effective learning.     

OCS: An effective caching scheme for video streaming on overlay networks

Video streaming is vital for many important applications such as distance learning, digital video libraries, and movie-on-demand. Since video streaming requires significant server and networking resources, caching has been used to reduce the demand on these resources. In this paper, we propose a novel collaboration scheme for video caching on overlay networks, called Overlay Caching Scheme (OCS), to further minimize service delays and loads placed on an overlay network for video streaming applications. OCS is not a centralized nor a hierarchical collaborative scheme. Despite its design simplicity, OCS effectively uses an aggregate storage space and capability of distributed overlay nodes to cache popular videos and serve nearby clients. Moreover, OCS is light-weight and adaptive to clients’ locations and request patterns. We also investigate other video caching techniques for overlay networks including both collaborative and non-collaborative ones. Compared with these techniques on topologies inspired from actual networks, OCS offers extremely low average service delays and approximately half the server load. OCS also offers smaller network load in most cases in our study.

Superiority of collaborative learning with complex tasks: A research note on an alternative affective explanation

Kirschner, Paas, and Kirschner (2009c) used the theoretical framework of cognitive load to explain why the learning of a group of collaborating individuals was more efficient than that of individuals learning alone with high-complexity tasks but not with low-complexity tasks. The authors argued that collaboration circumvented the limitations of an individual’s working memory by creating an expanded cognitive capacity and by allowing for the distribution of cognitive load among group members. Inspired by research on efficacy, this study explored an alternative affective explanation of the results. By measuring the amount of mental effort learners expected to invest in working on a learning task before actually carrying out the task, this study showed that learners who had to collaboratively solve a high-complexity problem expected to invest less mental effort than learners who had to solve the problem alone. When confronted with low-complexity tasks, the expected amount of mental effort did not differ.     

Towards Self-X cognitive manufacturing network: An industrial knowledge graph-based multi-agent reinforcement learning approach

Empowered by the advanced cognitive computing, industrial Internet-of-Things, and data analytics techniques, today’s smart manufacturing systems are ever-increasingly equipped with cognitive capabilities, towards an emerging Self-X cognitive manufacturing network with higher level of automation. Nevertheless, to our best knowledge, the readiness of ‘Self-X’ levels (e.g., self-configuration, self-optimization, and self-adjust/adaptive/healing) is still in the infant stage. To pave its way, this work stepwise introduces an industrial knowledge graph (IKG)-based multi-agent reinforcement learning (MARL) method for achieving the Self-X cognitive manufacturing network. Firstly, an IKG should be formulated based on the extracted empirical knowledge and recognized patterns in the manufacturing process, by exploiting the massive human-generated and machine-sensed multimodal data. Then, a proposed graph neural network-based embedding algorithm can be performed based on a comprehensive understanding of the established IKG, to achieve semantic-based self-configurable solution searching and task decomposition. Moreover, a MARL-enabled decentralized system is presented to self-optimize the manufacturing process, and to further complement the IKG towards Self-X cognitive manufacturing network. An illustrative example of multi-robot reaching task is conducted lastly to validate the feasibility of the proposed approach. As an explorative study, limitations and future perspectives are also highlighted to attract more open discussions and in-depth research for ever smarter manufacturing.     

Transductive Multi-Instance Multi-Label learning algorithm with application to automatic image annotation

Automatic image annotation has emerged as an important research topic due to its potential application on both image understanding and web image search. Due to the inherent ambiguity of image-label mapping and the scarcity of training examples, the annotation task has become a challenge to systematically develop robust annotation models with better performance. From the perspective of machine learning, the annotation task fits both multi-instance and multi-label learning framework due to the fact that an image is usually described by multiple semantic labels (keywords) and these labels are often highly related to respective regions rather than the entire image. In this paper, we propose an improved Transductive Multi-Instance Multi-Label (TMIML) learning framework, which aims at taking full advantage of both labeled and unlabeled data to address the annotation problem. The experiments over the well known Corel 5000 data set demonstrate that the proposed method is beneficial in the image annotation task and outperforms most existing image annotation algorithms.     

An examination of two mental workload measurement approaches to understanding multimedia learning

This study reports on an examination of two measures of mental workload: the NASA-TLX and Paas’ Subjective Cognitive Load (SCL) measure. The goal was to assess the relative efficacy of the measures in the design and research of multimedia learning environments. Benchmarks based on the literature as to the goals for mental workload measurement in learning research are established. A multifaceted study was conducted which manipulated various aspects of mental workload in order to study the utility of these two measures in detecting changes in load and their relationship to learning outcomes. The results indicate that the weighted version of the NASA-TLX provided little additional value over the unweighted version of the measure. While both the NASA-TLX and SCL measures were sensitive to changes in both intrinsic and extraneous load, the study revealed differences in the measures based on levels of each of these load factors. The study also concludes that a better understanding of the third factor, germane load, will be needed to both expand the theoretical framework about mental workload in instructional settings and further understand the utility of these two measures.     

An empirical study of Web browsing behaviour: Towards an effective Website design

Prior studies have suggested that a good Website design which facilitates a user’s Web browsing behaviour would generally lead to better user performance. In this research, we examine user Website behaviour as a way to understand Website design using a “think aloud” protocol analysis. Main theoretical contributions of this research are the illustration of the flow of cognitive processes during the Website browsing and the establishment of Website design dimensions – the meaning and content implied by Website content, its outward form, and the structure and navigation – in relation to user performance.     

Individual and group-based learning from complex cognitive tasks: Effects on retention and transfer efficiency

The effects of individual versus group learning (in triads) on efficiency of retention and transfer test performance in the domain of biology (heredity) among 70 high-school students were investigated. Applying cognitive load theory, the limitations of the working memory capacity at the individual level were considered an important reason to assign complex learning tasks to groups rather than to individuals. It was hypothesized that groups will have more processing capacity available for relating the information elements to each other and by doing so for constructing higher quality cognitive schemata than individuals if the high cognitive load imposed by complex learning tasks could be shared among group members. In contrast, it was expected that individuals who learn from carrying out the same complex tasks would need all available processing capacity for remembering the interrelated information elements, and, consequently, would not be able to allocate resources to working with them. This interaction hypothesis was confirmed by the data on efficiency of retention and transfer test performance; there was a favorable relationship between mental effort and retention test performance for the individual learners as opposed to a favorable relationship between transfer test performance and mental effort for the students who learned in groups.     

Automatic annotation of tennis games: An integration of audio,vision, and learning

Fully automatic annotation of tennis game using broadcast video is a task with a great potential but with enormous challenges. In this paper we describe our approach to this task, which integrates computer vision, machine listening, and machine learning. At the low level processing, we improve upon our previously proposed state-of-the-art tennis ball tracking algorithm and employ audio signal processing techniques to detect key events and construct features for classifying the events. At high level analysis, we model event classification as a sequence labelling problem, and investigate four machine learning techniques using simulated event sequences. Finally, we evaluate our proposed approach on three real world tennis games, and discuss the interplay between audio, vision and learning. To the best of our knowledge, our system is the only one that can annotate tennis game at such a detailed level.     

AVCD-FRA: A novel solution to automatic video cut detection using fuzzy-rule-based approach

Video shot boundary detection (SBD) is a fundamental step in automatic video content analysis toward video indexing, summarization and retrieval. Despite the beneficial previous works in the literature, reliable detection of video shots is still a challenging issue with many unsolved problems. In this paper, we focus on the problem of hard cut detection and propose an automatic algorithm in order to accurately determine abrupt transitions from video. We suggest a fuzzy rule-based scene cut identification approach in which a set of fuzzy rules are evaluated to detect cuts. The main advantage of the proposed method is that, we incorporate spatial and temporal features to describe video frames, and model cut situations according to temporal dependency of video frames as a set of fuzzy rules. Also, while existing cut detection algorithms are mainly threshold dependent; our method identifies cut transitions using a fuzzy logic which is more flexible. The proposed algorithm is evaluated on a variety of video sequences from different genres. Experimental results, in comparison with the most standard cut detection algorithms confirm our method is more robust to object and camera movements as well as illumination changes.     

Generic reflective feedback: An effective approach to developing clinical reasoning skills

Problem‐based learning can be an effective tool to develop clinical reasoning skills. However, it traditionally takes place in tutorial groups, giving students little flexibility in how and when they learn. This pilot study compared the effectiveness of generic reflective feedback (GRF) with tutorial‐based reflective feedback on the development of students' clinical reasoning skills. Case studies were placed in an interactive software package, with GRF for the nine patient cases (36 GRF forms in total). The GRF contained information about common mistakes and omissions made by students in previous years. By design, the GRF left some clinical questions unanswered in order to prompt student reflection, enhance the development of their clinical reasoning skills and give the 2010 students (n = 53) greater control over the way in which they engaged with the unit. Clinical reasoning skills development, assessed by responses to the 36 parts across the 9 cases and in the relevant section of the final exam, was compared with the 2009 tutorial cohort (n = 47 students). GRF enabled student progression through each of the nine online cases with minimal tutor intervention. The cohort receiving GRF performed as well as those students engaging in tutorials with individualized feedback. GRF was associated with a significant increase in student satisfaction and flexibility in how and when they learned. We conclude that GRF enabled online enhancement of a case‐based clinical reasoning unit. Further research on how GRF can optimize learning in different online settings is warranted.     

Adaptive leadership in trauma resuscitation teams: a grounded theory approach to video analysis

The detailed analysis of team interactions can be a source of insight into team processes and how teams interact with technology. Video recordings afford an exciting medium for such analysis. We describe a study of team leadership in the highly dynamic, high-stakes environment of trauma resuscitation. The study was conducted through video recording team activities in actual work settings and analysing the video data using a grounded theory approach. The primary research questions were: what are the functions of team leadership and how do they vary according to task situations? A corpus of 152 video segments from 18 trauma patient resuscitation cases was compiled to address these research questions. A catalog of team leadership functions was developed, along with a categorisation of the task situations in which team leadership occurred. The implications of this catalog and the mapping between leadership and task situations are discussed in relation to the findings from an interview study and a survey study on team leadership. The methodological advantages of a grounded theory approach for in-context video analysis for studying work are also discussed.     

Computer-Supported Collaborative Learning Requiring Immersive Presence (CSCLIP): An Introduction

The dramatic increase in distance learning (DL) enrollments in higher education is likely to continue. However, research on DL that includes psychomotor, cognitive, and affective skills is virtually non-existent. Indeed, DL for psychomotor skills has been viewed as impossible. Laboratory coursework, which we define as including the acquisition of psychomotor, cognitive, and affective skills, has become a limiting factor in the growth of DL. What is needed is a synergistic integration of technologies and Human-Computer Interface (HCI) principles from Computer Supported Collaborative Learning (CSCL), collaborative learning systems, and immersive presence technologies to enable achievement of psychomotor learning objectives. This paper defines the Computer Supported Collaborative Learning requiring Immersive Presence (CSCLIP) research area, and introduces this special issue of Information Systems Frontiers on CSCLIP and the six papers contained herein.This revised version was published online in March 2005 with corrections to the cover date     

Computer-supported collaborative learning with digital video cases in teacher education: The impact of teaching experience on knowledge convergence

Case-based learning is ascribed high potential with respect to the education of pre-service teachers as well as the further education of experienced in-service teachers, as it can provide opportunities for the application of professional knowledge to authentic classroom situations. In addition to the application and acquisition of knowledge, the aspect of knowledge convergence (i.e., increasing similarity of collaborative learners with respect to their knowledge) should be taken into account when homogeneous benefits are desired for all participants of a case-based training. A field study with 29 pre-service teachers and 24 in-service teachers from adult education was conducted to examine effects of different levels of teaching experience on knowledge application, knowledge acquisition, and knowledge convergence in a computer-supported collaborative case-based learning environment. An innovative approach to measuring knowledge convergence was introduced, using the positions of annotations (i.e., flags placed on the timeline of the case video) made by learners during their case analyses as an indicator of their shared focus of attention. This measure was found to positively correlate with the application and acquisition of professional knowledge. Teaching experience affected both processes and outcomes of case-based learning, suggesting different means of instructional support for pre-service and in-service teachers.     

VisionGo: Towards video retrieval with joint exploration of human and computer

This paper introduces an effective interactive video retrieval system named VisionGo. It jointly explores human and computer to accomplish video retrieval with high effectiveness and efficiency. It assists the interactive video retrieval process in different aspects: (1) it maximizes the interaction efficiency between human and computer by providing a user interface that supports highly effective user annotation and an intuitive visualization of retrieval results; (2) it employs a multiple feedback technique that assists users in choosing proper method to enhance relevance feedback performance; and (3) it facilitates users to assess the retrieval results of motion-related queries by using motion-icons instead of static keyframes. Experimental results based on over 160 h of news video shows demonstrate the effectiveness of the VisionGo system.     

Improving physicians' performance with a stroke CDSS: A cognitive fit design approach

A large body of evidence over many years suggests that clinical decision support systems (CDSSs) can be helpful in improving both clinical outcomes and adherence to evidence-based guidelines. CDSSs have however failed to show their effectiveness due to poor ease of use and integration within clinical workflows. This research therefore emphasizes a cognitive fit design approach to developing an effective CDSS to solve those issues. According to the cognitive fit theory CDSS should align with the problem and task representation in order to match the physician's mental model to reduce cognitive effort. Several guidelines based on the cognitive fit design are proposed. A stroke CDSS prototype following the design guidelines is developed to demonstrate its feasibility. The system usability test results showed that the developed stroke CDSS was acceptably accurate, able to lessen the cognitive effort as desired, and preferable for use due to the significant reduction in cognitive load. The developed artifact has shown the potential to benefit physicians. Finally, implications and conclusions are discussed.