Citation Context

...te the author features for each meme by taking the maximum, average, and standard deviation among the group of authors who have posted or reposted the meme by t1. We use Support Vector Machines (SVM) =-=[7]-=- to predict meme importance using each of the volume, static, and dynamic network features above as well as their combination. 7. EXPERIMENTS AND OBSERVATIONS In this section, we present our experimen...

Citation Context

...ion and/or a fee. MM’11, November 28–December 1, 2011, Scottsdale, Arizona, USA. Copyright 2011 ACM 978-1-4503-0616-4/11/11 ...$10.00. text-based information networks such as Twitter rely on retweets =-=[5, 18]-=-, hashtags, mentions, or trackbacks to identify influence and trending topics [1], similar functions for large video-sharing repository is lacking. A reliable videobased“quote”tracking and popularity ...

Citation Context

...lysis in social event streams. In Box A, the border color of meme clusters denotes the event they are from: green (top): Iran; orange (bottom): SwineFlu. tors who play notable roles in social changes =-=[14]-=-. We use features derived from the meme network model to predict the lifespan of memes, achieving an area-under-ROC-curve (AUC) measure of 0.78. The main contributions of this work are as follows: • W...

Citation Context

...l pairs of frames that are within a given distance threshold has complexity O(N2), while ANN indexing can speed this up significantly. We use the Fast Library for Approximate Nearest Neighbor (FLANN) =-=[21]-=- to implement the indexing structure and ANN lookup. FLANN automatically selects the best indexing structure between k-means tree and kd-tree, and chooses the appropriate tree parameters for a given d...

Citation Context

...ion and/or a fee. MM’11, November 28–December 1, 2011, Scottsdale, Arizona, USA. Copyright 2011 ACM 978-1-4503-0616-4/11/11 ...$10.00. text-based information networks such as Twitter rely on retweets =-=[5, 18]-=-, hashtags, mentions, or trackbacks to identify influence and trending topics [1], similar functions for large video-sharing repository is lacking. A reliable videobased“quote”tracking and popularity ...

Citation Context

...sions. We train SVM classifiers [7] by searching over hyperparameters and different kernel types—linear, polynomial, and radial basis function. We measure the classification performance using the AUC =-=[11]-=- and plot the results in Figure 11 (top). We can see that author connectivity alone, or fusing all features together, are stronger features than content volume alone. Using all three types of features...

Citation Context

...ive research areas in both multimedia analysis and social media mining. YouTube has been the focal platform for many social network monitoring studies. The first large-scale YouTube measurement study =-=[6]-=- characterized content category distributions, and tracked exact duplicates of popular videos. Benevenuto et al. studied video response actions on YouTube using metadata [3], and De Choudhury et al. m...

Citation Context

... micro-blogs [5, 18], where users often quote the original text message verbatim, having little freedom for remixing and context changes within the 140 character limit. Another example is MemeTracker =-=[19]-=-, which tracks the lifecycles of popular phrases among blogs and news websites. Prior studies have shown that the frequency of video reuse can be used as an implicit video quality indicator [23]. Howe...

Citation Context

...oudhury et al. monitored user comments to determine interesting conversations [10]. Recently, early views of YouTube videos have been used to predict ultimate popularity, characterized by view counts =-=[25]-=-. Quoting, duplication, and reposting are popular phenomena in online information networks. One well-known example is retweeting on micro-blogs [5, 18], where users often quote the original text messa...

Citation Context

...spect ratio; performing de-noising; and applying contrast-limited histogram equalization to correct for contrast and gamma differences. We use a frame similarity metric based on the color correlogram =-=[16]-=- that captures the local spatial correlation of pairs of colors. The color correlogram is rotation-, scale-, and to some extent, viewpoint-invariant. It was designed to tolerate moderate changes in ap...

Citation Context

...ng by keyframes is suitable for capturing community video remixing and provides a good balance of speed and accuracy in public evaluations on video copy detection tasks [22]. Hashing-based techniques =-=[27]-=- are another alternative for speeding up ANN queries but their precision is typically not more than 50%, which is too low for reliably tracking all sets of near-duplicates in large collections. The AN...

Citation Context

...f Figure 4 Box D. Therefore, we perform transitive closure on the neighbor relationship to find full equivalence classes of near-duplicate sets. This is done with an efficient set unionfind algorithm =-=[12]-=- that runs in amortized time of O(E), where E is the number of matched pairs. 5.3 Discussion Our design choices for the visual meme detection system aim to find the best combination of accuracy and sp...

Citation Context

...been on userdependent definitions of duplicates [8], speeding up detection on image sequence, frame, or local image points [26], and scaling out to web-scale computations using large compute clusters =-=[20]-=-. We note, however, that most prior work in this area is concerned with optimizing retrieval accuracy of detecting near-duplicate frames or sequences, rather than tracking large-scale duplication beha...

Citation Context

... YouTube measurement study [6] characterized content category distributions, and tracked exact duplicates of popular videos. Benevenuto et al. studied video response actions on YouTube using metadata =-=[3]-=-, and De Choudhury et al. monitored user comments to determine interesting conversations [10]. Recently, early views of YouTube videos have been used to predict ultimate popularity, characterized by v...

Citation Context

...ct duplicates of popular videos. Benevenuto et al. studied video response actions on YouTube using metadata [3], and De Choudhury et al. monitored user comments to determine interesting conversations =-=[10]-=-. Recently, early views of YouTube videos have been used to predict ultimate popularity, characterized by view counts [25]. Quoting, duplication, and reposting are popular phenomena in online informat...

Citation Context

...ifference 4tji. The first model is insensitive to 4tji, so it can accommodate the resurgence of popular memes, as seen in textual memes [19]. The power law decay comes from known behaviors on YouTube =-=[9]-=-, and it also agrees with our observations on the recency of the content returned by the YouTube search API. ω∗ij = νij (i, j) ∈ EG (1) ω′ij = νij4t−ηji (2) The unit for time t is in days. We estimate...

Citation Context

...g near-duplicates in images and video has been a problem of interest since the early years of content-based retrieval. Recent focus of this problem has been on userdependent definitions of duplicates =-=[8]-=-, speeding up detection on image sequence, frame, or local image points [26], and scaling out to web-scale computations using large compute clusters [20]. We note, however, that most prior work in thi...

Citation Context

... most prior work in this area is concerned with optimizing retrieval accuracy of detecting near-duplicate frames or sequences, rather than tracking large-scale duplication behavior. Kennedy and Chang =-=[17]-=- tracked editing and provenance of images on the web, with a focus on distinguishing different types of image edits and their ideological perspective. Our work, in comparison, tracks large-scale video...

Citation Context

...acker [19], which tracks the lifecycles of popular phrases among blogs and news websites. Prior studies have shown that the frequency of video reuse can be used as an implicit video quality indicator =-=[23]-=-. However, none of the prior work has defined the unit for retweet or meme on a video sharing network. Tracking near-duplicates in images and video has been a problem of interest since the early years...

Citation Context

...EMIXES Visual memes are defined as frequently reposted video segments or images. It has been observed that users tend to create “curated selections based on what they liked or thought was important” (=-=[24]-=-, page 270). News event collections are particularly suited for studying large-scale user curation, since remixing is more prevalent here than on video genres designed for self-expression, such as vid...

Citation Context

...the early years of content-based retrieval. Recent focus of this problem has been on userdependent definitions of duplicates [8], speeding up detection on image sequence, frame, or local image points =-=[26]-=-, and scaling out to web-scale computations using large compute clusters [20]. We note, however, that most prior work in this area is concerned with optimizing retrieval accuracy of detecting near-dup...

Citation Context

...oth content and metadata such as authorship and creation time, and focuses on inferring social roles in video propagation. Several recent works have looked at YouTube phenomena— Biel and Gatica-Perez =-=[4]-=- focused on individual social behavior such as non-verbal cues, while Hong et al. [15] presented content summarization by monitoring a query over time. In comparison, we use visual memes to capture th...

Citation Context

...ng social roles in video propagation. Several recent works have looked at YouTube phenomena— Biel and Gatica-Perez [4] focused on individual social behavior such as non-verbal cues, while Hong et al. =-=[15]-=- presented content summarization by monitoring a query over time. In comparison, we use visual memes to capture the behavior of large groups and to track information dissemination. 3. VISUAL MEMES AND...

Citation Context

...trast, gamma), and video production edits (such as the superimposition of text, captions, borders, transition effects). Most of these are well-known as the targets of visual copy detection benchmarks =-=[22]-=-. In this paper, meme refers both to individual 54 !"#$%&'& !"#$%&(& )$*$& +,%-& $./*01$+& !" !" 2%3-34$&& !"#$%&0/5$& !" #" $!" $#" %!" &'$('!)" &'%!'!)" &'%*'!)" *'+'!)" *'$$'!)" *'$,'!)" *'%#'!)" ,...

Citation Context

... ease of publishing and sharing videos has outpaced the progress of modern search engines, collaborative tagging sites, and content aggregation services—leaving users to deal with a deluge of content =-=[2]-=-. This information overload problem is particularly prominent for linear media (such as audio, video, animations), where at-a-glance impressions are hard to develop and are often unreliable. While *Ar...

Citation Context

...or of overall influence. This is an example of the very unequal distribution of views that characterizes this domain. To quantify the inequality of views-counts, we have computed the Gini coefficient =-=[13]-=- of this data set, and find it to have the extreme value of 0.94, both for videos with or without nearduplicates. The Gini coefficient, used in economics, ranges from 0 (each video with an equal numbe...