Explore the vast range of titles available.

This collection departs from the observation that online forms of communication—the email, blog, text message, tweet—are actually haunted by old epistolary forms: the letter and the diary. By examining the omnipresence of writing across a variety of media, the collection adds the category of Epistolary Screens to genres of self-expression, both literary (letters, diaries, auto-biographies) and screenic (romance dramas, intercultural cinema, essay films, artists’ videos and online media). The category Epistolary encapsulates an increasingly paradoxical relation between writing and the self: first, it describes selves that are written in graphic detail via letters, diaries, blogs, texts, emails and tweets; second, it acknowledges that absence complicates communication, bringing people together in an entangled rather than ordered way. The collection concerns itself with the changing visual/textual texture of screen media and examines what is at stake for our understanding of self-expression when it takes Epistolary forms.

This collection departs from the observation that online forms of communication - the email, blog, text message, tweet - are actually haunted by old epistolary forms: the letter and the diary. By examining the omnipresence of writing across a variety of media, the collection adds the category of Epistolary Screens to genres of self-expression, both literary (letters, diaries, auto-biographies) and screenic (romance dramas, intercultural cinema, essay films, artists' videos and online media). The category Epistolary encapsulates an increasingly paradoxical relation between writing and the self: first, it describes selves that are written in graphic detail via letters, diaries, blogs, texts, emails and tweets; second, it acknowledges that absence complicates communication, bringing people together in an entangled rather than ordered way. The collection concerns itself with the changing visual/textual texture of screen media and examines what is at stake for our understanding of self-expression when it takes Epistolary forms.

The marine environment presents a unique set of challenges for human–robot interaction. Communicating with gestures is a common way for interacting between the diver and autonomous underwater vehicles (AUVs). However, underwater gesture recognition is a challenging visual task for AUVs due to light refraction and wavelength color attenuation issues. Current gesture recognition methods classify the whole image directly or locate the hand position first and then classify the hand features. Among these purely visual approaches, textual information is largely ignored. This paper proposes a visual–textual model for underwater hand gesture recognition (VT-UHGR). The VT-UHGR model encodes the underwater diver’s image as visual features, the category text as textual features, and generates visual–textual features through multimodal interactions. We guide AUVs to use image–text matching for learning and inference. The proposed method achieves better performance than most existing purely visual methods on the dataset CADDY, demonstrating the effectiveness of using textual patterns for underwater gesture recognition.

Searching persons in large-scale image databases with the query of natural language description is a more practical and important application in video surveillance. Intuitively, for person search, the core issue should be the visual–textual association, which is still an extremely challenging task, due to the contradiction between the high abstraction of textual description and the intuitive expression of visual images. In this paper, aim for more consistent visual–textual features and better inter-class discriminate ability, we propose a text-based person search approach with visual–textual attention on the hardest and semi-hard negative pairs mining. First, for the visual and textual attentions, we designed a Smoothed Global Maximum Pooling (SGMP) to extract more concentrated visual features, and also the memory attention based on LSTM’s cell unit for more strictly correspondence matching. Second, while we only have labeled positive pairs, more valuable negative pairs are mined by defining the cross-modality-based hardest and semi-hard negative pairs. After that, we combine the triplet loss on the single modality with the hardest negative pairs, and the cross-entropy loss on cross-modalities with both the hardest and semi-hard negative pairs, to train the whole network. Finally, to evaluate the effectiveness and feasibility of the proposed approach, we conduct extensive experiments on the typical person search dataset: CUHK-PEDES, in which our approach achieves satisfactory performance, e.g, the top1 accuracy of 55.32 % . Besides, we also evaluate the semi-hard pair mining method in the COCO caption dataset and validate its effectiveness and complementary.

Aspect-level multimodal sentiment analysis aims to ascertain the sentiment polarity of a given aspect from a text review and its accompanying image. Despite substantial progress made by existing research, aspect-level multimodal sentiment analysis still faces several challenges: (1) Inconsistency in feature granularity between the text and image modalities poses difficulties in capturing corresponding visual representations of aspect words. This inconsistency may introduce irrelevant or redundant information, thereby causing noise and interference in sentiment analysis. (2) Traditional aspect-level sentiment analysis predominantly relies on the fusion of semantic and syntactic information to determine the sentiment polarity of a given aspect. However, introducing image modality necessitates addressing the semantic gap in jointly understanding sentiment features in different modalities. To address these challenges, a multi-granularity visual-textual feature fusion model (MG-VTFM) is proposed to enable deep sentiment interactions among semantic, syntactic, and image information. First, the model introduces a multi-granularity hierarchical graph attention network that controls the granularity of semantic units interacting with images through constituent tree. This network extracts image sentiment information relevant to the specific granularity, reduces noise from images and ensures sentiment relevance in single-granularity cross-modal interactions. Building upon this, a multilayered graph attention module is employed to accomplish multi-granularity sentiment fusion, ranging from fine to coarse. Furthermore, a progressive multimodal attention fusion mechanism is introduced to maximize the extraction of abstract sentiment information from images. Lastly, a mapping mechanism is proposed to align cross-modal information based on aspect words, unifying semantic spaces across different modalities. Our model demonstrates excellent overall performance on two datasets.

Due to increased migration and global aging, transnational caregiving plays an increasingly significant role in supporting work-family integration in Canadian society. Yet, there is limited research exploring racialized transnational carer employees’ (R TCEs’) experiences in Canada. TCEs are immigrants working in paid employment in Canada and providing unpaid care to family and/or friends across nations. This unpaid care can include emotional, physical and/or financial support. The data for this article were drawn from a larger study that examined R TCEs’ experience using arts-based and qualitative inquiry. Seventeen participants (male = 10, female = 7, other = 0) provided an art piece (e.g. poem, artifact, photograph, and drawing) as well as a written or verbal description of their piece’s meaning. This paper applies a semiotic framework and “Systematic Visual-Textual Analysis” to triangulate our analysis of participant art pieces and the meaning they gave to these creative products. Our analysis illustrates the multi-dimensional experience of transnational carer employees in Canada, through the common and overlapping symbolism of transition, care, love, and motivation. The research provides a cross-cultural, nuanced, and wholistic perspective on transnational care by R TCEs in Canada, while taking a novel analytical approach that allows for the systemic application of semiotics to arts-based analysis. Our findings have the potential to inform the implementation and content of caregiving supports in Canadian workplaces, post-secondary institutions, and medical care, as well as the application of semiotics and systematic visual-textual analysis in social science.

Background Sentiment analysis (SA) has turned out to be a new pattern in social networking, avidly helping people to realize views expressed in user-generated content and conventional platforms of social media. For performing numerous social media analytics tasks, SA of online user-produced content is vital. The performance of the sentiment classifiers utilizing a single modality, i.e., visual or textual, is still not matured because of the wide variety of data platforms. Methods In this paper, we propose a new framework called VIsual-TExtual SA (VITESA) that carries out visual analysis and textual analysis for polarity classification. In the VITESA framework, Brownian Movement-based Meerkat Clan Algorithm-centered DenseNet (BMMCA-DenseNet) is proposed that integrates textual and visual information for robust SA. In the visual phase, the images that are in the Flickr dataset are taken as input, and the operations: (1) preprocessing (2) feature extraction, and (3) feature selection utilizing Improved Coyote Optimization Algorithm (ICOA) are executed. In the textual phase, the user comments as of the Twitter dataset are taken as input, and the operations: (1) preprocessing (2) word embedding using adaptive Embedding for Language Models (ELMo), (3) emoticon and non-emoticon feature extraction, and SentiWordNet polarity assignment is carried out. The final stages of both phases are given as input to the proposed BMMCA-DenseNet classifier and intended to categorize the data into positive and negative polarity. The performance of BMMCA-DenseNet is compared with certain existing algorithms, and various performance metrics are evaluated. Results The proposed BMMCA-DenseNet classifier performs the polarity classification of the visual-textual data into two classes: positive or negative. The classifier categorizes the polarity of visual-textual data comprising 97% of accuracy, 94.44% of precision, 94.41% of recall, 94.41% of F-measure, 91.75% of Matthew’s Correlation Coefficient, 94.43% of sensitivity, 97.13% of specificity, and also minimal error. Conclusions The experiment is performed to evaluate the performance of the proposed method. The outcomes exhibit that BMMCA-DenseNet attains remarkable performance over other existing techniques. The result enhances the textual-visual communication systematically to improve sentiment prediction utilizing both information sources.

Our two main research aims were to elicit children’s liking preferences for forty-five different vegetables by asking them which vegetables they have eaten and which were their two favorite vegetables. Our third aim was to disseminate our Child Vegetable Liking Survey for broader use. Children (n = 448) in the first, second, third, and fourth grades completed the prototype Child Vegetable Liking Survey. Students were read a paper-and-pencil demographic survey that asked questions about their age, gender, and race ethnicity, followed by a checklist of forty-five vegetables that they have ever eaten and a prompt to circle their two favorite vegetables. A color photograph supported the textual (word) list of forty-five vegetables to help students with recognition of their vegetable liking preferences. Implications for survey development with children are discussed in the context of functional health literacy and interactive health literacy. A list of expository texts with vegetable photographs are included for researchers to share with educators who desire to build background knowledge about a variety of vegetables with picture book compatibility. Conclusions: The top three vegetables with the highest frequencies and percentages “ever eaten” by the children included corn (n = 418; 93.3%), broccoli (n = 394; 97.9%), and carrots (n = 380; 84.8%), followed by potatoes, celery, lettuce, tomato, peas, sweet potato, and cucumber. A high number of students reported NOT ever eating bok choy (n = 413; 92.2%), yellow beans (n =404; 90.2%), kohlrabi (n = 399; 89.1%), followed by okra, rhubarb, parsnips, rutabaga, collard greens, yams, edamame, and artichokes. Corn was chosen as the favorite vegetable by the most students (n = 147; 34.6%). The second and third favorite vegetables were carrots (n = 88; 20.8%) and broccoli (n = 77; 18.2%), followed by potatoes, cucumber, pumpkin, sweet potato, red pepper, avocado, lettuce, and tomato. The following vegetables were never chosen by the children as one of their two favorite vegetables: collard greens, kohlrabi, parsnip, rutabaga, turnips, and yellow beans.

The rapid popularization of various online media services have attracted large amounts of consumers and shown us a large potential market of video advertising. In this paper, we propose interactive service recommendation based on ad concept hierarchy and contextual search. Instead of traditional ODP (Open Directory Project) based approach, we built a ad domain based concept hierarchy to make the most of the product details over the e-commerce sites. Firstly, we capture the summarization images related to the advertising product in the video content and search visually similar product images from the built product image database. Then, we aggregate the visual tags and textual tags with K-line clustering. Finally, we map them to the product concept space and make keywords suggestion, and users can interactively select keyframes or keywords to personalize their intentions by textual re-search. Experiments and comparison show that the system can accurately provide effective advertising suggestions.