Visual Sentiment Prediction with Deep Convolutional Neural Networks

Xu, Can; Cetintas, Suleyman; Lee, Kuang-Chih; Li, Lijia

doi:10.48550/arxiv.1411.5731

Cited by 30 publications

(40 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the early studies, a CNN is often directly used as the off-the-shelf tool without any modification. For example, Xu et al [30] trained two classifiers following the two fully connected (FC) layers (FC7 and FC8) of an existing basic network (AlexNet), respectively. The experimental results show that the classifier after the FC7 (0.649) layer performs better than that after the FC8 (0.615).…”

Section: Learning-based Methodsmentioning

confidence: 99%

“…Quantitative Comparison of Representative Deep Methods. As shown in Table 8, we conduct experiments to fairly compare four representative learning-based methods, including DCNN [30], RCA [107], WSCNet [12], and PDANet [111]. We replace the original backbone with four different backbones to evaluate the effectiveness and robustness, including AlexNet [136], VGG-16 [114], ResNet-50 [18], and Inception-v3 [137].…”

Section: Learning-based Methodsmentioning

confidence: 99%

“…Since then, several representative hand-crafted features have been designed, such as the low-level artistic elements [8], mid-level artistic principles [9], and highlevel Adjective Noun Pairs (ANPs) [10]. In 2014, transfer learning is conducted from a CNN in which parameters are pre-trained by large-scale data [30]. To tackle the subjectivity challenge, both personalized emotion prediction [31,15] and emotion distribution learning [14,32,33,34] are considered.…”

Section: Brief Historymentioning

confidence: 99%

“…The newly trained deep model named DeepSentiBank [103] performs better than non-deep Sen-tiBank on sentiment prediction. Benefiting from transfer learning, Xu et al [30] transferred the parameters of a CNN trained on the large-scale dataset (ImageNet) to the task of predicting sentiments. They extracted a 4096-dimensional representation from the fully connected (FC) layer FC7 and a 1000-dimensional representation from FC8, respectively, as the image-level features.…”

Section: Global Featuresmentioning

confidence: 99%

See 3 more Smart Citations

Affective Image Content Analysis: Two Decades Review and New Perspectives

Zhao

Yao

Yang

et al. 2022

IEEE Trans. Pattern Anal. Mach. Intell.

View full text Add to dashboard Cite

Images can convey rich semantics and induce various emotions in viewers. Recently, with the rapid advancement of emotional intelligence and the explosive growth of visual data, extensive research efforts have been dedicated to affective image content analysis (AICA). In this survey, we will comprehensively review the development of AICA in the recent two decades, especially focusing on the state-of-the-art methods with respect to three main challenges -the affective gap, perception subjectivity, and label noise and absence. We begin with an introduction to the key emotion representation models that have been widely employed in AICA and description of available datasets for performing evaluation with quantitative comparison of label noise and dataset bias. We then summarize and compare the representative approaches on (1) emotion feature extraction, including both handcrafted and deep features, (2) learning methods on dominant emotion recognition, personalized emotion prediction, emotion distribution learning, and learning from noisy data or few labels, and (3) AICA based applications. Finally, we discuss some challenges and promising research directions in the future, such as image content and context understanding, group emotion clustering, and viewer-image interaction.

show abstract

Section: Learning-based Methodsmentioning

confidence: 99%

Section: Learning-based Methodsmentioning

confidence: 99%

Section: Brief Historymentioning

confidence: 99%

Section: Global Featuresmentioning

confidence: 99%

See 2 more Smart Citations

Affective Image Content Analysis: Two Decades Review and New Perspectives

Zhao

Yao

Yang

et al. 2022

IEEE Trans. Pattern Anal. Mach. Intell.

View full text Add to dashboard Cite

show abstract

“…Global features are directly extracted from the whole images. One direct and intuitive method is to employ the output of the last few fully connected (FC) layers as deep features, using either pretrained or finetuned CNN models [59,4,69]. The last few FC layers correspond to high-level semantic features, which might be not enough to represent emotions, especially for abstract images.…”

Section: Deep Featuresmentioning

confidence: 99%

Computational Emotion Analysis From Images: Recent Advances and Future Directions

Zhao¹,

Huang²,

Tang³

et al. 2021

Preprint

View full text Add to dashboard Cite

Emotions are usually evoked in humans by images. Recently, extensive research efforts have been dedicated to understanding the emotions of images. In this chapter, we aim to introduce image emotion analysis (IEA) from a computational perspective with the focus on summarizing recent advances and suggesting future directions. We begin with commonly used emotion representation models from psychology. We then define the key computational problems that the researchers have been trying to solve and provide supervised frameworks that are generally used for different IEA tasks. After the introduction of major challenges in IEA, we present some representative methods on emotion feature extraction, supervised classifier learning, and domain adaptation. Furthermore, we introduce available datasets for evaluation and summarize some main results. Finally, we discuss some open questions and future directions that researchers can pursue.

show abstract

Negative-Supervised Cascaded Deep Learning for Traffic Sign Classification

Xie

Yang

et al. 2015

Communications in Computer and Information Science

View full text Add to dashboard Cite

In this paper, we propose a novel deep learning framework for object classification called negative-supervised cascaded deep learning. There are two hierarchies in our cascaded method: the first one is a convolutional neural network trained on positive-only samples, which is used to select supervisory samples from a negative library. The second one is inherited from the trained first CNN. It is trained on positive and negative samples, which are selected from domain related database by utilizing negative-supervised mechanism. Experiments are applied this idea to traffic sign classification using two classic convolutional neural networks, LeNet-5 and AlexNet as baselines. Classification rates improved by 0.7% and 1.1% with LeNet-5 and AlexNet respectively, which demonstrates the efficiency and superiority of our proposed framework.

show abstract

Visual Sentiment Prediction with Deep Convolutional Neural Networks

Cited by 30 publications

References 13 publications

Affective Image Content Analysis: Two Decades Review and New Perspectives

Affective Image Content Analysis: Two Decades Review and New Perspectives

Computational Emotion Analysis From Images: Recent Advances and Future Directions

Negative-Supervised Cascaded Deep Learning for Traffic Sign Classification

Contact Info

Product

Resources

About