scholarly journals Topic-Oriented Text Features Can Match Visual Deep Models of Video Memorability

2021 ◽  
Vol 11 (16) ◽  
pp. 7406
Author(s):  
Ricardo Kleinlein ◽  
Cristina Luna-Jiménez ◽  
David Arias-Cuadrado ◽  
Javier Ferreiros ◽  
Fernando Fernández-Martínez
Keyword(s):  
Linear Regression ◽  
Natural Language ◽  
Video Clip ◽  
Visual Media ◽  
Topic Detection ◽  
Detection Model ◽  
Visual Models ◽  
Text Features ◽  
Semantic Dependencies ◽  
Source Of Information

Not every visual media production is equally retained in memory. Recent studies have shown that the elements of an image, as well as their mutual semantic dependencies, provide a strong clue as to whether a video clip will be recalled on a second viewing or not. We believe that short textual descriptions encapsulate most of these relationships among the elements of a video, and thus they represent a rich yet concise source of information to tackle the problem of media memorability prediction. In this paper, we deepen the study of short captions as a means to convey in natural language the visual semantics of a video. We propose to use vector embeddings from a pretrained SBERT topic detection model with no adaptation as input features to a linear regression model, showing that, from such a representation, simpler algorithms can outperform deep visual models. Our results suggest that text descriptions expressed in natural language might be effective in embodying the visual semantics required to model video memorability.

scholarly journals Efficient natural language classification algorithm for detecting duplicate unsupervised features

2021 ◽  
Vol 20 (3) ◽  
pp. 623-653
Author(s):  
Saud Altaf ◽  
Sofia Iqbal ◽  
Muhammad Waseem Soomro
Keyword(s):  
Natural Language ◽  
Short Term Memory ◽  
Short Term ◽  
Vocabulary Size ◽  
Language Understanding ◽  
Inverse Document Frequency ◽  
Classification Technique ◽  
Document Frequency ◽  
Text Features ◽  
Long Short Term Memory

This paper focuses on capturing the meaning of Natural Language Understanding (NLU) text features to detect the duplicate unsupervised features. The NLU features are compared with lexical approaches to prove the suitable classification technique. The transfer-learning approach is utilized to train the extraction of features on the Semantic Textual Similarity (STS) task. All features are evaluated with two types of datasets that belong to Bosch bug and Wikipedia article reports. This study aims to structure the recent research efforts by comparing NLU concepts for featuring semantics of text and applying it to IR. The main contribution of this paper is a comparative study of semantic similarity measurements. The experimental results demonstrate the Term Frequency–Inverse Document Frequency (TF-IDF) feature results on both datasets with reasonable vocabulary size. It indicates that the Bidirectional Long Short Term Memory (BiLSTM) can learn the structure of a sentence to improve the classification.

scholarly journals A Compositional Framework for Grounding Language Inference, Generation, and Acquisition in Video

2015 ◽  
Vol 52 ◽  
pp. 601-713 ◽  
Author(s):  
Haonan Yu ◽  
N. Siddharth ◽  
Andrei Barbu ◽  
Jeffrey Mark Siskind
Keyword(s):  
Natural Language ◽  
Video Clip ◽  
Spatial Relations ◽  
Inference Generation ◽  
Video Clips ◽  
The Third

We present an approach to simultaneously reasoning about a video clip and an entire natural-language sentence. The compositional nature of language is exploited to construct models which represent the meanings of entire sentences composed out of the meanings of the words in those sentences mediated by a grammar that encodes the predicate-argument relations. We demonstrate that these models faithfully represent the meanings of sentences and are sensitive to how the roles played by participants (nouns), their characteristics (adjectives), the actions performed (verbs), the manner of such actions (adverbs), and changing spatial relations between participants (prepositions) affect the meaning of a sentence and how it is grounded in video. We exploit this methodology in three ways. In the first, a video clip along with a sentence are taken as input and the participants in the event described by the sentence are highlighted, even when the clip depicts multiple similar simultaneous events. In the second, a video clip is taken as input without a sentence and a sentence is generated that describes an event in that clip. In the third, a corpus of video clips is paired with sentences which describe some of the events in those clips and the meanings of the words in those sentences are learned. We learn these meanings without needing to specify which attribute of the video clips each word in a given sentence refers to. The learned meaning representations are shown to be intelligible to humans.

scholarly journals Audio-visual media to improve sexual-reproduction health knowledge among adolescent

2020 ◽  
Vol 9 (1) ◽  
pp. 138
Author(s):  
Sitti Nur Djannah ◽  
Sulistyawati Sulistyawati ◽  
Tri Wahyuni Sukesi ◽  
Surahma Asti Mulasari ◽  
Fatwa Tentama
Keyword(s):  
Sexual Reproduction ◽  
Junior High School ◽  
Health Knowledge ◽  
Visual Media ◽  
Wilcoxon Test ◽  
Post Intervention ◽  
Before And After ◽  
The Mean ◽  
The Difference ◽  
Source Of Information

<span>Lacking knowledge among adolescents affects their understanding of some problems related to sexual-reproduction health. Electronic media recognized as the favored source of information for adolescents. This research aimed to assess the effect of audio-visual media to the increasing of sexual-reproduction knowledge. We conducted a before and after without control informal experimental study design into 153 students in the 1st-3rd grade of junior high school. The effect of the intervention was assessed through the difference between pre- and post-intervention by using the Wilcoxon test. The mean score of the respondent pre and post-intervention was significantly increasing. The audiovisual increased the knowledge of the adolescent regarding sexual-reproduction health</span>

scholarly journals TRACKING DAN CHROMAKEY SEBAGAI ELEMEN TEKNIK DESAIN EFEK VISUAL Studi Kasus: Efek Visual Video Klip

2020 ◽  
Vol 5 (1) ◽  
pp. 85
Author(s):  
Ardiyan Ardiyan ◽  
Satria Mahardika ◽  
Melki Sadekh Mansuan ◽  
Veronica Wijayanti
Keyword(s):  
Video Clip ◽  
Visual Media ◽  
Visual Effect ◽  
Alpha Channel ◽  
Early Production ◽  
Study Case ◽  
Technical Factors ◽  
Green Screen ◽  
E Mail

<p><strong>Abstrak</strong></p><p>Tracking and Chromakey as Visual Effect Design Technique (Study Case: Video Clip Visual Effect. Visual effect in audio visual media and animation is commonly use, especially chromakey technique or usually described as green screen or blue screen, a certain color element to acknowledge the alpha channel. In compositing technique, the using of tracking and chromakey technique are both known as visual effect technique. But on online editing works, the result seldom not at its best. This problem is usually occurred from early production process until the final process in the process of video making. The quality of final result based on early raw footage used. The objective of this research is to know the technical factors to achieve better method appliance in order to enhance the quality of final result. The study case is video clip that has short duration, more variative and has natural compositing result. The literature studies of software and user approached is the method use to analyses the problem. The result of this research is the understanding of tracking and chromakey method in video clip making process to achieve the visual effect needed. </p><p> </p><p><strong> Abstrak</strong></p><p>Tracking dan Chromakey Sebagai Elemen Teknik Desain Efek Visual (Studi Kasus: Efek Visual Video Klip). Efek visual dalam media audio visual maupun animasi sudah sangat lazim digunakan, khususnya penggunaan teknik chromakey atau yang lebih umum disebut green screen atau blue screen, yaitu penggunaan elemen warna tertentu untuk menentukan alpha channel. Dalam lingkup teknik compositing, penggunaan teknik tracking dan chromakey ini dapat digolongkan sebagai teknik efek visual. Dalam pengerjaan online editing, hasil akhir terkadang terasa kurang maksimal apabila ditinjau dari hasil kedua teknik ini. Hal ini dapat diartikan adanya permasalahan yang selalu terkait dengan pengolahan produksi awal sampai akhir dalam pengerjaan sebuah video. Kualitas hasil akhir akan ditentukan dari raw footage awal yang akan digunakan. Penelitian ini bertujuan untuk mengetahui faktor teknis yang ada, sehingga lebih baik dalam penggunaan metode yang dilakukan serta akan meningkatkan kualitas hasil akhir. Sebagai studi kasus adalah video klip dengan mempertimbangkan durasi yang tidak lama, lebih variatif dan mengarah dalam pendekatan hasil compositing yang natural. Selain itu metode studi literatur tentang perangkat lunak juga dilakukan, sehingga pendekatan metode dan pola pikir pengguna perangkat lunak lebih memahami dalam melakukan kedua proses ini. Hasil dari penelitian ini adalah pemahaman terhadap metode tracking dan chromakey dalam pembuatan video klip untuk menghasilkan efek visual yang dibutuhkan. Kata kunci: efek visual, compositing, video klip, tracking, chromakey<br />*) Jurusan Desain Komunikasi Visual, Program Animasi School Of Design Universitas Bina Nusantara e-mail: [email protected]</p>

scholarly journals Content Noise Detection Model Using Deep Learning in Web Forums

2020 ◽  
Vol 12 (12) ◽  
pp. 5074
Author(s):  
Jiyoung Woo ◽  
Jaeseok Yun
Keyword(s):  
Machine Learning ◽  
Deep Learning ◽  
Learning Model ◽  
Detection Model ◽  
Proposed Model ◽  
Web Forum ◽  
Web Forums ◽  
Conventional Machine ◽  
Text Features ◽  
Deep Learning Model

Spam posts in web forum discussions cause user inconvenience and lower the value of the web forum as an open source of user opinion. In this regard, as the importance of a web post is evaluated in terms of the number of involved authors, noise distorts the analysis results by adding unnecessary data to the opinion analysis. Here, in this work, an automatic detection model for spam posts in web forums using both conventional machine learning and deep learning is proposed. To automatically differentiate between normal posts and spam, evaluators were asked to recognize spam posts in advance. To construct the machine learning-based model, text features from posted content using text mining techniques from the perspective of linguistics were extracted, and supervised learning was performed to distinguish content noise from normal posts. For the deep learning model, raw text including and excluding special characters was utilized. A comparison analysis on deep neural networks using the two different recurrent neural network (RNN) models of the simple RNN and long short-term memory (LSTM) network was also performed. Furthermore, the proposed model was applied to two web forums. The experimental results indicate that the deep learning model affords significant improvements over the accuracy of conventional machine learning associated with text features. The accuracy of the proposed model using LSTM reaches 98.56%, and the precision and recall of the noise class reach 99% and 99.53%, respectively.

scholarly journals From WOM to aWOM – the evolution of unpaid influence: a perspective article

2019 ◽  
Vol 75 (1) ◽  
pp. 314-318 ◽  
Author(s):  
Nigel L. Williams ◽  
Nicole Ferdinand ◽  
John Bustard
Keyword(s):  
Decision Making ◽  
Natural Language Processing ◽  
Natural Language ◽  
Language Processing ◽  
Distribution Systems ◽  
Word Of Mouth ◽  
Customer Engagement ◽  
Content Type ◽  
Dominant Source ◽  
Source Of Information

Purpose Advances in artificial intelligence (AI) natural language processing may see the emergence of algorithmic word of mouth (aWOM), content created and shared by automated tools. As AI tools improve, aWOM will increase in volume and sophistication, displacing eWOM as an influence on customer decision-making. The purpose of this paper is to provide an overview of the socio technological trends that have encouraged the evolution of informal infulence strategies from WOM to aWOM. Design/methodology/approach This paper examines the origins and path of development of influential customer communications from word of mouth (WOM) to electronic word of mouth (eWOM) and the emerging trend of aWOM. The growth of aWOM is theorized as a result of new developments in AI natural language processing tools along with autonomous distribution systems in the form of software robots and virtual assistants. Findings aWOM may become a dominant source of information for tourists, as it can support multimodal delivery of useful contextual information. Individuals, organizations and social media platforms will have to ensure that aWOM is developed and deployed responsibly and ethically. Practical implications aWOM may emerge as the dominant source of information for tourist decision-making, displacing WOM or eWOM. aWOM may also impact online opinion leaders, as they may be challenged by algorithmically generated content. aWOM tools may also generate content using sensors on personal devices, creating privacy and information security concerns if users did not give permission for such activities. Originality/value This paper is the first to theorize the emergence of aWOM as autonomous AI communication within the framework of unpaid influence or WOM. As customer engagement will increasingly occur in algorithmic environments that comprise person–machine interactions, aWOM will influence future tourism research and practice.

scholarly journals Image Caption Generation Model Based on Object Detector

2018 ◽  
pp. 5-14
Author(s):  
Aghasi Poghosyan ◽  
Hakob Sarukhanyan
Keyword(s):  
Natural Language ◽  
Object Detection ◽  
Information Extraction ◽  
Semantic Information ◽  
Generation Model ◽  
Single Model ◽  
Detection Model ◽  
Generator Performance ◽  
Image Caption Generation ◽  
Image Caption

Automated semantic information extraction from the image is a difficult task. There are works, which can extract image caption or object names and their coordinates. This work presents object detection and automated caption generation implemented via a single model. We have built an image caption generation model on top of object detection model. We have added extra layers on object detector to increase caption generator performance. We have developed a single model that can detect objects, localize them and generate image caption via natural language.

scholarly journals Webshell Detection Based on Executable Data Characteristics of PHP Code

2021 ◽  
Vol 2021 ◽  
pp. 1-12
Author(s):  
Zulie Pan ◽  
Yuanchao Chen ◽  
Yu Chen ◽  
Yi Shen ◽  
Xuanzhen Guo
Keyword(s):  
State Of The Art ◽  
Recall Rate ◽  
Remote Access ◽  
Detection Accuracy ◽  
Controlled Experiments ◽  
Data Set ◽  
Detection Model ◽  
Proposed Model ◽  
Text Features ◽  
And Control

A webshell is a malicious backdoor that allows remote access and control to a web server by executing arbitrary commands. The wide use of obfuscation and encryption technologies has greatly increased the difficulty of webshell detection. To this end, we propose a novel webshell detection model leveraging the grammatical features extracted from the PHP code. The key idea is to combine the executable data characteristics of the PHP code with static text features for webshell classification. To verify the proposed model, we construct a cleaned data set of webshell consisting of 2,917 samples from 17 webshell collection projects and conduct extensive experiments. We have designed three sets of controlled experiments, the results of which show that the accuracy of the three algorithms has reached more than 99.40%, the highest reached 99.66%, the recall rate has been increased by at least 1.8%, the most increased by 6.75%, and the F1 value has increased by 2.02% on average. It not only confirms the efficiency of the grammatical features in webshell detection but also shows that our system significantly outperforms several state-of-the-art rivals in terms of detection accuracy and recall rate.

scholarly journals THE INFLUENCE OF AUDIO-VISUAL MEDIA ON STUDENTS 'LEARNING MOTIVATION DURING THE PANDEMIC TIME OF COVID-19 AT MI AL FAJAR PRINGSEWU

2021 ◽  
Vol 1 (02) ◽  
pp. 55-60
Author(s):  
Muhtarom ◽  
Evi Gusliana ◽  
Syeh Al Ngarifin ◽  
Moh. Masrur
Keyword(s):  
Linear Regression ◽  
Student Learning ◽  
Student Motivation ◽  
Visual Media ◽  
Learning Motivation ◽  
Education Student ◽  
Motivation To Learn ◽  
Simple Linear Regression ◽  
Regression Test ◽  
The World

Abstract The Covid 19 outbreak has an impact on various areas of life, especially the world of education. Student learning is required to transform and adapt to the conditions of the 4.0 century. This is what motivates researchers to conduct research on children's learning motivation during the Covid-19 pandemic. The aim of this research is to determine the effect of using audio-visual media on student motivation during the Covid-19 pandemic in grade IV MI Al Fajar Pringsewu. This research is a quantitative study with data collection techniques using a questionnaire. The data were analyzed using simple linear regression test. The results of this study indicate there is an effect of audio visual media on student learning motivation during the Covid-19 pandemic with a known value tcount > ttable (4.118 > 2.056) then Ho was rejected and Ha accepted. The big influence of audio-visual media on student learning motivation namely 57,3 % Keywords : Covid-19, audio-visual media, motivation to learn

Sign in / Sign up

Export Citation Format

Share Document