Fast Text Caption Localization on Video Using Visual Rhythm

Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications

Leite

2011

Text detection in images has been studied and improved for decades. There are many works that extend the existing methods for analyzing videos, however, few of them create or adapt approaches that consider inherent characteristics of videos, such as temporal information. This work proposes a very fast method for identifying video frames that contain text through a special data structure called visual rhythm. The method is robust to detect video captions with respect to font styles, color intensity, and text orientation. A data set was built in our experiments to compare and evaluate the effectiveness of the proposed method.

Section: Resultsmentioning

confidence: 98%

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Fast Rotation-Invariant Video Caption Detection Based on Visual Rhythm

Valio

Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications

Leite

2011

“…Visual rhythm is a simplification of a video content in a 2D image obtained by sampling regions of the video. Applications of this concept can be found in the work by Chun et al [33] that use visual rhythms for fast text caption localization on video, and Guimarães et al [34] who propose a method for gradual transition detection in videos. The use of visual rhythm in our work is crucial since it allows us to capture patterns that are present in the Fourier Spectrum Videos providing an effective way of viewing a video as a still image.…”

Section: Calculation Of the Visual Rhythmsmentioning

confidence: 98%

Using Visual Rhythms for Detecting Video-Based Facial Spoof Attacks

Pinto

Schwartz

IEEE Trans.Inform.Forensic Secur.

et al. 2015

116

Spoofing attacks or impersonation can be easily accomplished in a facial biometric system wherein users without access privileges attempt to authenticate themselves as valid users, in which an impostor needs only a photograph or a video with facial information of a legitimate user. Even with recent advances in biometrics, information forensics and security, vulnerability of facial biometric systems against spoofing attacks is still an open problem. Even though several methods have been proposed for photo-based spoofing attack detection, attacks performed with videos have been vastly overlooked, which hinders the use of the facial biometric systems in modern applications. In this paper, we present an algorithm for video-based spoofing attack detection through the analysis of global information which is invariant to content, since we discard video contents and analyze contentindependent noise signatures present in the video related to the unique acquisition processes. Our approach takes advantage of noise signatures generated by the recaptured video to distinguish between fake and valid access videos. For that, we use the Fourier spectrum followed by the computation of video visual rhythms and the extraction of different characterization methods. For evaluation, we consider the novel Unicamp Video-Attack Database (UVAD) which comprises 17,076 videos composed of real access and spoofing attack videos. In addition, we evaluate the proposed method using the Replay-Attack Database, which contains photo-based and video-based face spoofing attacks.

“…The application of the visual rhythm can be found on the works including the one by Chun et al [26] for fast text caption localization on videos and by Guimarães et al [27] for identifying cut and gradual transitions on videos.…”

Section: A Video Characterizationmentioning

confidence: 99%

Video-Based Face Spoofing Detection through Visual Rhythm Analysis

Pinto

2012 25th SIBGRAPI Conference on Graphics, Patterns and Images

Schwartz

et al. 2012

Fig. 1. Given a training set consisting of videos of valid accesses, video-based spoofs, and a video for testing, first we extract a noise signature of every video (training and testing) and calculate the Fourier Spectrum on logarithmic scale for each video frame. Thereafter, we create visual rhythms for each video and train a machine learning classifier using either the pixel intensities directly as features or a summarized version of the visual rhythms using gray level co-occurrence matrices. With a trained classifier, we are able to test a visual rhythm for a given video under investigation and point out whether it is a valid access or a video-based spoof.Abstract-Recent advances on biometrics, information forensics, and security have improved the accuracy of biometric systems, mainly those based on facial information. However, an ever-growing challenge is the vulnerability of such systems to impostor attacks, in which users without access privileges try to authenticate themselves as valid users. In this work, we present a solution to video-based face spoofing to biometric systems. Such type of attack is characterized by presenting a video of a real user to the biometric system. To the best of our knowledge, this is the first attempt of dealing with video-based face spoofing based in the analysis of global information that is invariant to video content. Our approach takes advantage of noise signatures generated by the recaptured video to distinguish between fake and valid access. To capture the noise and obtain a compact representation, we use the Fourier spectrum followed by the computation of the visual rhythm and extraction of the graylevel co-occurrence matrices, used as feature descriptors. Results show the effectiveness of the proposed approach to distinguish between valid and fake users for video-based spoofing with nearperfect classification results.