Efficiency and usability study of innovative computer-aided transcription strategies for video lecture repositories

Abstract

ElsevierValor Miró, JD.; Silvestre Cerdà, JA.; Civera Saiz, J.; Turró Ribalta, C.; Juan Císcar, A. (2015). Efficiency and usability study of innovative computer-aided transcription strategies for video lecture repositories. AbstractVideo lectures are widely used in education to support and complement face-to-face lectures. However, the utility of these audiovisual assets could be further improved by adding subtitles that can be exploited to incorporate added-value functionalities such as searchability, accessibility, translatability, note-taking, and discovery of content-related videos, among others. Today, automatic subtitles are prone to error, and need to be reviewed and post-edited in order to ensure that what students see on-screen are of an acceptable quality. This work investigates different user interface design strategies for this post-editing task to discover the best way to incorporate automatic transcription technologies into large educational video repositories. Our three-phase study involved lecturers from the Universitat Politècnica de València (UPV) with videos available on the poliMedia video lecture repository, which is currently over 10,000 video objects. Simply by conventional post-editing automatic transcriptions users almost reduced to half the time that would require to generate the transcription from scratch. As expected, this study revealed that the time spent by lecturers reviewing automatic transcriptions correlated directly with the accuracy of said transcriptions. However, it is also shown that the average time required to perform each individual editing operation could be precisely derived and could be applied in the definition of a user model. In addition, the second phase of this study presents a transcription review strategy based on confidence measures (CM) and compares it to the conventional post-editing strategy. Finally, a third strategy resulting from the combination of that based on CM with massive adaptation techniques for ASR achieved to improve the transcription review efficiency in comparison with the two aforementioned strategies.

Transcribing against time

“…We first compared the computational efficiency of segmenting according to the algorithm proposed in §6, as opposed to the baseline method of formulating the problem as an integer linear program and using an off-the-shelve ILP solver. As our solver, we used GUROBI, which is highly optimized commercial software and among the fastest ILP solvers available, 9 while our Java implementation of the proposed algorithm is straightforward, with little code level optimization. We used the initial CM to segment increasingly large subsets of our data.…”

“…[21] and [22] use a more practical approach that specifies a constrained optimization problem by allowing only a limited time budget for supervision, similar to our approach. Note that works on annotation apart from the active learning community have usually assumed stopping criterions based on confidence thresholds [3,4,9], which measure only relative improvement and may not be intuitive to configure in practice.…”

“…Our solutions to these challenges are as follows: (1) We propose to limit the transcription time budget to a fixed value, in order to reflect one's desired cost-quality tradeoff. How to choose a suitable time budget is task specific, but we argue that it is more practical than configuring a confidence threshold as was required in some previous works [3,4,9]. Moreover, by periodically updating segmentations ( §4) we can recover from inaccurately predicted correction times, making sure that the budget is not over-or underspent.…”

Impact of Deep Learning on Localizing and Recognizing Handwritten Text in Lecture Videos

“…In the MOOCs, it is vital to understand the lecture videos for educational research as it has become synonymous with distance learning. The better understanding of the lecture video lies in the vital cues like the figures, images and text [8] [9] [10] [11]. Among these vital cues, the text is available in almost all lectures as it can be utilized for variety of tasks like the extracting class notes, generation of keywords, search enabling and video indexing.…”

Multimodal Classification of Teaching Activities from University Lecture Recordings

“…Our model is independent of the automated transcription tool and can be used with the transcriptions provided by any service such as YouTube, Apple's Siri, Zoom video communications or others. In our case, we used the open-source MLLP transcription and translation platform [39,40], which turned out to translate scientific and technical terms better than YouTube.…”