Abstract
Passive content fingerprinting is widely used for video content identification and monitoring. However, many challenges remain unsolved especially for partial-copies detection. The main challenge is to find the right balance between the computational cost of fingerprint extraction and fingerprint dimension, without compromising detection performance against various attacks (robustness). Fast video detection performance is desirable in several modern applications, for instance, in those where video detection involves the use of large video databases or in applications requiring real-time video detection of partial copies, a process whose difficulty increases when videos suffer severe transformations. In this context, conventional fingerprinting methods are not fully suitable to cope with the attacks and transformations mentioned before, either because the robustness of these methods is not enough or because their execution time is very high, where the time bottleneck is commonly found in the fingerprint extraction and matching operations. Motivated by these issues, in this work we propose a content fingerprinting method based on the extraction of a set of independent binary global and local fingerprints. Although these features are robust against common video transformations, their combination is more discriminant against severe video transformations such as signal processing attacks, geometric transformations and temporal and spatial desynchronization. Additionally, we use an efficient multilevel filtering system accelerating the processes of fingerprint extraction and matching. This multilevel filtering system helps to rapidly identify potential similar video copies upon which the fingerprint process is carried out only, thus saving computational time. We tested with datasets of real copied videos, and the results show how our method outperforms state-of-the-art methods regarding detection scores. Furthermore, the granularity of our method makes it suitable for partial-copy detection; that is, by processing only short segments of 1 second length.
Highlights
Nowadays, video-based applications flooded the Internet traffic
That method achieves better results against many attacks using long video segments (10s-200s), it is very susceptible to simple attacks that change the color correlation. Another method based on global features was presented in [21], where the Discrete Cosine Transform (DCT) coefficients obtained from temporary informative representative images (TIRI) are used as the fingerprint
We have presented an efficient content fingerprinting method based on the extraction of a set of independent binary global and local fingerprints
Summary
Video-based applications flooded the Internet traffic. According to the Cisco Visual Networking Index [1], by 2019 it is projected that globally, Internet Protocol (IP) video traffic will be 80% of all IP traffic. The film industry identifies camcording as a major problem, and industry representatives estimate that camcords are responsible for at least 90% percent of the first available versions of illegally distributed new release films [8] For those reasons, camcording and edition attacks are considered as the most common visual attacks. Video monitoring and copy detection systems are based mainly on passive content fingerprinting and watermarking [11, 12]. Proposed, these are not able to achieve the detection combining the desirable characteristics of a) robustness against the intentional attacks in real cases (mentioned above), b) effectiveness in the detection of short partial copies [10] and c) efficiency for real-time applications [14]. Aiming at dealing with these issues, we propose using a set of visual fingerprints whose combination leverages the performance in the partial-video copy detection task, outperforming the results obtained by state-of-the-art methods.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have