English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Despite the rapid progress of the techniques for image classification, videoannotation has remained a challenging task. Automated video annotation would bea breakthrough technology, enabling users to search within the videos.Recently, Google introduced the Cloud Video Intelligence API for videoanalysis. As per the website, the system can be used to "separate signal fromnoise, by retrieving relevant information at the video, shot or per frame"level. A demonstration website has been also launched, which allows anyone toselect a video for annotation. The API then detects the video labels (objectswithin the video) as well as shot labels (description of the video events overtime). In this paper, we examine the usability of the Google's Cloud VideoIntelligence API in adversarial environments. In particular, we investigatewhether an adversary can subtly manipulate a video in such a way that the APIwill return only the adversary-desired labels. For this, we select an image,which is different from the video content, and insert it, periodically and at avery low rate, into the video. We found that if we insert one image every twoseconds, the API is deceived into annotating the video as if it only containedthe inserted image. Note that the modification to the video is hardlynoticeable as, for instance, for a typical frame rate of 25, we insert only oneimage per 50 video frames. We also found that, by inserting one image persecond, all the shot labels returned by the API are related to the insertedimage. We perform the experiments on the sample videos provided by the APIdemonstration website and show that our attack is successful with differentvideos and images.

Deceiving Google's Cloud Video Intelligence API Built for Summarizing Videos