Survey on visual sentiment analysis

Ortis, Alessandro; Farinella, Giovanni Maria; Battiato, Sebastiano

doi:10.1049/iet-ipr.2019.1270

Cited by 74 publications

(27 citation statements)

References 102 publications

(195 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Rao(c)’s method [ 3 ] obtains local areas of different scales through faster R-CNN based on FPN which requires a tedious pre-training process that the entire model needs to be pre-trained on the COCO [ 44 ] and EmotionROI datasets to obtain object perception capabilities. However, our model does not require other additional cumbersome pre-training processes.…”

Section: Methodsmentioning

confidence: 99%

“…Compared with the non-emotional stimulus content in the image, the affective content attracts the attention of the viewer more strongly, and the viewer has a more detailed understanding of the affective stimulus content [ 2 ]. Therefore, the purpose of visual sentiment analysis is to understand the emotional impact of visual materials on viewers [ 3 ], which plays an important role in opinion mining, user behavior prediction, emotional image retrieval, game scene modeling and other aspects.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Multi-Level Context Pyramid Network for Visual Sentiment Analysis

Qing

et al. 2021

Sensors

View full text Add to dashboard Cite

Sharing our feelings through content with images and short videos is one main way of expression on social networks. Visual content can affect people’s emotions, which makes the task of analyzing the sentimental information of visual content more and more concerned. Most of the current methods focus on how to improve the local emotional representations to get better performance of sentiment analysis and ignore the problem of how to perceive objects of different scales and different emotional intensity in complex scenes. In this paper, based on the alterable scale and multi-level local regional emotional affinity analysis under the global perspective, we propose a multi-level context pyramid network (MCPNet) for visual sentiment analysis by combining local and global representations to improve the classification performance. Firstly, Resnet101 is employed as backbone to obtain multi-level emotional representation representing different degrees of semantic information and detailed information. Next, the multi-scale adaptive context modules (MACM) are proposed to learn the sentiment correlation degree of different regions for different scale in the image, and to extract the multi-scale context features for each level deep representation. Finally, different levels of context features are combined to obtain the multi-cue sentimental feature for image sentiment classification. Extensive experimental results on seven commonly used visual sentiment datasets illustrate that our method outperforms the state-of-the-art methods, especially the accuracy on the FI dataset exceeds 90%.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Multi-Level Context Pyramid Network for Visual Sentiment Analysis

Qing

et al. 2021

Sensors

View full text Add to dashboard Cite

show abstract

“…A final suggestion for future research is to utilize Visual Sentiment Analysis to investigate how text messages and images between partners might trigger emotions and relate to relationship quality and satisfaction. 26…”

Section: Limitations and Future Directionsmentioning

confidence: 99%

Moderation of Technology Use in the Association Between Self-Isolation During COVID-19 Pandemic and Adolescents' Romantic Relationship Quality

Wright

Wachs

2021

Cyberpsychology, Behavior, and Social Networking

View full text Add to dashboard Cite

The purpose of this study was to examine the moderating effects of technology use for relationship maintenance on the longitudinal associations among self-isolation during the coronavirus-19 (COVID-19) pandemic and romantic relationship quality among adolescents. Participants were 239 (120 female; M age = 16.69, standard deviation [SD] = 0.61; 60 percent Caucasian) 11th and 12th graders from three midwestern high schools.To qualify for this study, adolescents had to be in the same romantic relationship for the duration of the study, *7 months (M length of relationship = 10.03 months). Data were collected in October of 2019 (Time 1) and again 7 months later in May of 2020 (Time 2). Adolescents completed a romantic relationship questionnaire at Time 1 and again at Time 2, along with questionnaires on frequency of self-isolation during the COVID-19 pandemic and use of technology for romantic relationship maintenance. Findings revealed that increases in selfisolation during the COVID-19 pandemic related positively to the use of technology for romantic relationship maintenance and negatively to Time 2 romantic relationship quality. High use of technology for romantic relationship maintenance buffered against the negative effects of self-isolation during the COVID-19 pandemic on adolescents' romantic relationship quality 7 months later, whereas low use strengthened the negative relationship between self-isolation during the COVID-19 pandemic and romantic relationship quality. These findings suggest the importance of considering the implications of societal crisis or pandemics on adolescents' close relationships, particularly their romantic relationships.

show abstract

“…Figure 1 describes the taxonomy of sentiment analysis. The literature regarding speculation of sentiments from image data is rather limited [ 17 ]. Moreover, being the latest and challenging task, there is a paucity of public datasets which makes it harder to construct a benchmark that can lay the foundation of a firm state of the art.…”

Section: Related Workmentioning

confidence: 99%

Human Sentiment and Activity Recognition in Disaster Situations Using Social Media Images Based on Deep Learning

Sadiq

Ahn

Choi

2020

Sensors

View full text Add to dashboard Cite

A rapidly increasing growth of social networks and the propensity of users to communicate their physical activities, thoughts, expressions, and viewpoints in text, visual, and audio material have opened up new possibilities and opportunities in sentiment and activity analysis. Although sentiment and activity analysis of text streams has been extensively studied in the literature, it is relatively recent yet challenging to evaluate sentiment and physical activities together from visuals such as photographs and videos. This paper emphasizes human sentiment in a socially crucial field, namely social media disaster/catastrophe analysis, with associated physical activity analysis. We suggest multi-tagging sentiment and associated activity analyzer fused with a a deep human count tracker, a pragmatic technique for multiple object tracking, and count in occluded circumstances with a reduced number of identity switches in disaster-related videos and images. A crowd-sourcing study has been conducted to analyze and annotate human activity and sentiments towards natural disasters and related images in social networks. The crowdsourcing study outcome into a large-scale benchmark dataset with three annotations sets each resolves distinct tasks. The presented analysis and dataset will anchor a baseline for future research in the domain. We believe that the proposed system will contribute to more viable communities by benefiting different stakeholders, such as news broadcasters, emergency relief organizations, and the public in general.

show abstract

Survey on visual sentiment analysis

Cited by 74 publications

References 102 publications

Multi-Level Context Pyramid Network for Visual Sentiment Analysis

Multi-Level Context Pyramid Network for Visual Sentiment Analysis

Moderation of Technology Use in the Association Between Self-Isolation During COVID-19 Pandemic and Adolescents' Romantic Relationship Quality

Human Sentiment and Activity Recognition in Disaster Situations Using Social Media Images Based on Deep Learning

Contact Info

Product

Resources

About