[PDF] Understanding image virality | Semantic Scholar (2024)

A novel pairwise reformulation of the virality prediction problem as an attribute prediction task is presented and a novel algorithm to model image virality on online media using a pairwise neural network is developed.

This paper studies the prediction of popular and viral image diffusion in Pinterest and shows that image meta and poster's information are strong predictors for predicting popular image cascades while imageMeta and initial propagation patterns are useful to predict viral image cascading.

This work proposes Diffusion-LSTM, a memory-based deep recurrent network that learns to recursively predict the entire diffusion path of an image through a social network and demonstrates its capability of generating diffusion trees, and shows that the generated trees closely resemble ground-truth trees.

A probabilistic method to generate massive popularity-discriminable image pairs is described, based on which the first large-scale image database for intrinsic IPA (I$^2$PA) is established and a psychophysical experiment is conducted to analyze various aspects of human behavior in I$^1$PA.

A tree-structured long short-term memory (LSTM) network that learns and predicts the entire diffusion path of an image in a social network by combining user social features and image features together with the encoded diffusion path history stored in an explicit memory cell.

This work introduces a pooling layer that learns the size of the support area to be averaged: the learned top-N average (LENA) pooling, and assesses the effectiveness of the LENA layer by appending it on top of a convolutional siamese architecture and evaluates its performance on the task of predicting and localizing virality.

This work presents a novel algorithm to model image virality on online networks using the increasingly popular deep convolutional neural network architectures and provides significant insights into the features that are responsible for promoting virality.

This project has introduced a new dataset from Twitter for studying virality and defined an annotation score using Twitter metadata and trained a deep learning Siamese model to predict virality of individual posts using relative virality in pairs of posts.

This paper re-visits the tweet popularity prediction problem by considering all data modalities: tweet language semantics, embedded images, author' social relationships, and the diffusion process of tweets to model the content of tweets, and proposes a joint-embedding neural network that combines visual, textual, and social cues together.

...

...

The importance of image cues, such as color, gradients, deep learning features and the set of objects present, as well as the importance of various social cues such as number of friends or number of photos uploaded that lead to high or low popularity of images are shown.

Several virality phenomena that emerge when taking into account visual characteristics of images (such as orientation, mean saturation, etc.) are described.

This work presents a method built upon psychological theories and web mining to automatically construct a large-scale Visual Sentiment Ontology (VSO) consisting of more than 3,000 Adjective Noun Pairs (ANP) and proposes SentiBank, a novel visual concept detector library that can be used to detect the presence of 1,200 ANPs in an image.

This paper demonstrates a simple, yet powerful method to automatically select high aesthetic quality images from large image collections and demonstrates that an aesthetics classifier trained on describable attributes can provide a significant improvement over baseline methods for predicting human quality judgments.

This paper statistically study the correlations among popular memes and their wordings, and proposes a robust nonparanormal model to learn the stochastic dependencies among the image, the candidate descriptions, and the popular votes.

It is found that, amongst content features, URLs and hashtags have strong relationships with retweetability and the number of followers and followees as well as the age of the account seem to affect retweetability, while, interestingly, thenumber of past tweets does not predict retweetability of a user's tweet.

The approach finds a set of nodes that guarantees spreading to the entire network under the tipping model and often finds seed sets that are several orders of magnitude smaller than the population size and outperform nodal centrality measures in most cases.

This paper explores how a number of factors relate to human perception of importance using what people describe as a proxy for importance, and builds models to predict what will be described about an image given either known image content, or image content estimated automatically by recognition systems.

This work proposes a generative model over the joint space of attribute ranking outputs, and proposes a novel form of zero-shot learning in which the supervisor relates the unseen object category to previously seen objects via attributes (for example, ‘bears are furrier than giraffes’).

...

...