Overview

Digital videos contain rich information and evidence about an event. These videos can easily be tampered with user-friendly video editing tools like Adobe Premier, After Effect, GNU Gimp, and Vegas etc., and uploaded on the social media for propaganda and malicious purposes. It also reduces the credibility of videos on the social media and in the court of law. Such issues have made forensic analysis essential to ensure the authenticity and integrity of videos. 

Background

The video forgery detection methods can broadly be categorized into two groups: active and passive. Passive techniques do not require embedded information (watermark, digital signature) unlike active techniques, and hence can be applied to authenticate any video. Videos can be forged by three different ways: (i) spatial or object-based tampering, (ii) temporal or frame based tampering, and (iii) spatio-temporal tampering. An example of which is shown in Figure below. The object in red rectangle is present in original video Figure (a) but this object is deleted from the frames to tamper the video as shown in Figure(b). In spatial forgery, the actual information is concealed by deleting an object from different frames and viewers are misguided by providing them false information. The purpose of this type of forgery is not merely retouching or changing format, but to hide the facts for propaganda or other criminal intentions. As such, this type of forgery is dangerous and has a negative impact on the society. 

Applications

The applications of this research in many fields of life such as media groups, insurance companies, marriage bureaus, investigation agencies, social media, and courts of law.

Expected Results

The outcome of the research is providing a system which can detect the video forgery, and would be able to localize the forgery as well. 

Data sets

Although there were limited data sets available to test the result of the research. The Data sets for forgery are as under. 

• O. I. Al-Sanjary, A. A. Ahmed, and G. Sulong, "Development of a video tampering dataset for forensic investigation", Forensic science international, vol. 266, pp. 565-572, 2016.

• G. Qadir, S. Yahaya, and A. T. Ho, "Surrey university library for forensic analysis (SULFA) of video content", in Image Processing (IPR 2012), IET Conference on, London, UK, 2012, pp. 1-6, doi:10.1049/cp.2012.0422

• E. Ardizzone and G. Mazzola, "A Tool to Support the Creation of Datasets of Tampered Videos," in Image Analysis and Processing— ICIAP 2015, ed: Springer, 2015, pp. 665-675., doi:10.1007/978-3- 319-23234-8_61

• C.-C. Hsu, T.-Y. Hung, C.-W. Lin, and C.-T. Hsu, "Video forgery detection using correlation of noise residue", in Multimedia Signal Processing, 2008 IEEE 10th Workshop on, Queensland, Australia, 2008, pp. 170-174, doi:10.1109/mmsp.2008.4665069 

• P. Bestagini, S. Milani, M. Tagliasacchi, and S. Tubaro, "Local tampering detection in video sequences", in Multimedia Signal Processing (MMSP), 2013 IEEE 15th International Workshop on, Pula (CA), Italy, 2013, pp. 488-493, doi:10.1109/MMSP.2013.6659337 

Related Publications

1. M. Saddique, K. Asghar, U. I. Bajwa, M. Hussain, Z. Habib "Spatial Video Forgery Detection and Localization using Texture Analysis of Consecutive Frames," Advances in Electrical and Computer Engineering, vol.19, no.3, pp.97-108, 2019, doi:10.4316/AECE.2019.03012. Impact Factor: 0.65 (External link) 

2. Mubbashar Saddique, Khurshid Asghar, Tariq Mehmood, Muhammad Hussain, Zulfiqar Habib, “Robust Video Content Authentication using Video Binary Pattern and Extreme Learning Machine”, in IJACSA Volume 10 (8), PP. 264-269, ISI Indexed (External link) 

3. Khurshid Asghar, Mubbashar Sadddique, Inam ul Haq, M. Ahmad Nawaz-ul-Ghani, Ghulam Ali, “Stacked Support Vector Machine Ensembles for Cross-Culture Emotions Classification” in International Journal of Computer Science and Network Security, Vol. 19 , No. 7, pp. 23-30, 2019. ISI Indexed (External link) 

4. Khurshid Asghar, Xianfang Sun, Paul L. Rosin, Mubbashar Saddique, Muhammad Hussain, Zulfiqar Habib (2019), "Edge-Texture Feature based Image Forgery Detection with Cross Dataset Evaluation", Machine Vision and Applications, pp 1-20, 2019. Impact Factor: 1.788 (External link)