Authors
Oinam Olivia Chanu, Riya Bansal
Abstract
In the subject of study known as media forensics, many types of media, including audio, video, and digital pictures, are analysed and investigated using scientific methodologies and procedures. Finding evidence that can be utilised in court or other investigations is the main goal of media forensic. Experts in media forensics evaluate media files using a variety of tools and methods, such as data recovery, picture and audio enhancement, and authentication. In this paper we are going to trace the origin and properties of the files uploaded in different social media platforms. This paper situates such as framework and suggests a novel approach to determine whether a picture originates from a social network and, more specifically, to determine which image has been downloaded. The method is predicated on the idea that each social network employs an odd and mostly unknown alteration that yet leaves some recognisable traces on the image, and that these traces may be retrieved to showcase every site. The proposed technique successfully distinguishes distinct social network sources by using trained classifiers. Experimental findings on same image shot from different social media cameras and finding the differences in the pictures under varied operational circumstances attest to the feasibility of such a distinction. Keywords: Enhancement, Authentication, Retrieved, Recognisable, Feasibility.
Introduction
The collection, analysis, interpretation and presentation of audio, video, image evidence gathered during investigations and legal proceedings is known as media forensics (www.artsandmedia.ucdenver.edu). Social media is very important in today’s digital environment and in daily life. Everyday numerous and various types of images, videos are uploaded in the social media platform every single second. However, we are not very sure about its properties and origin of the file, whether they are being manipulated or genuine. To study about this matter, media forensics is one of the greatest solutions to these uncertainties. It seems as though one cannot live without a social networking site. Growing worries about the reliability of digital media were caused in the previous ten to twelve years by the dissemination of simple editing tools that were available to a larger public. In this case, the problem has lately been made worse by the creation of new classes of artificial intelligence algorithms that enable people to create high-quality fake photos and videos (like Deepfakes) without the need for any specialised technical knowledge.
Additionally, multimedia material plays a crucial part in the digital lives of people and civilizations, substantially contributing to the viral transmission of information through social media and web channels. As a result, our society can no longer ignore the need to create tools to maintain the reliability of images and videos shared on social media and web platforms (Pasquini et al., 2021).
According to research, 1.81 trillion photographs are shot worldwide each year, and 6.9 billion of them are shared on WhatsApp daily. 1.3 billion photos are shared daily on Instagram, with around 100 million appearing in posts and more than 1 billion appearing in stories and conversations. 3.8 billion images are being shared everyday through snapchat, 2.1 billion in Facebook and 1 million images through Flickr (www.photutorial.com). One-third of the population is able to access the internet and post photos to websites and social media. These data transmit a number of other things due to their digital characters. Details of their life history, such as the originating device and any processing they have undergone. When visual evidence is used in a crime, this information could become important. Multimedia forensics has been suggested as a possible remedy for this situation in order to examine photographs and videos in order to learn more about their life cycle. All these years, a number of methods are created to evaluate digital images, concentrating on problems with identifying the source device and judging the veracity of the material (Shullani et al., 2017).
An image may be taken and uploaded simultaneously to one or more social networks due to the increasing usage of smartphones. On the flip side, unlawful actions are mushrooming by abusing such digital stuff to accomplish different, occasionally ignoble, goals. Facing these increasing problems, related to different social media in our day today life, it is of great importance to know the origin and properties of images, videos uploaded in social media. By knowing the origin of the image like provenance of the image, from which camera it was shot, to which social media it was uploaded first, it would be of considerable assistance in media forensics, law enforcement and intelligence services in finding out the culprits responsible for creating misleading visual contents, manipulated and misused the images. More generally, it can assist in preserving the credibility of digital media and reducing the effects of misinformation by enforcing trustable sources by finding the properties details like checking out the hash value, metadata details, Exif information of the image, etc (Caldelli et al., 2017).
References
ByMatic Broz | March 10, 2023 (2023) How many photos are there? (2023) 50+ photos statistics, Photutorial. https://photutorial.com/photos-statistics/ (Accessed: April 4, 2023).
Caldelli, Roberto, et al. “Image Origin Classification Based on Social Network Provenance.” IEEE Transactions on Information Forensics and Security, vol. 12, no. 6, Institute of Electrical and Electronics Engineers, June 2017, pp. 1299–308. https://doi.org/10.1109/tifs.2017.2656842.
College of Arts & Media (no date) About The National Center for Media Forensics. https://artsandmedia.ucdenver.edu/areas-of-study/national-center-for-media-forensics/about-the-national-center-for-media-forensics#:~:text=Media%20forensics%20is%20scientific%20study%20into%20the%20collection%2C,during%20the%20course%20of%20investigations%20and%20litigious%20proceedings. (Accessed: April 4, 2023).
Hasinoff, Samuel W., et al. “Burst Photography for High Dynamic Range and Low-light Imaging on Mobile Cameras.” ACM Transactions on Graphics, vol. 35, no. 6, Association for Computing Machinery, Nov. 2016, pp. 1–12. https://doi.org/10.1145/2980179.2980254.
Pasquini, Cecilia, et al. “Media Forensics on Social Media Platforms: A Survey.” EURASIP Journal on Information Security, vol. 2021, no. 1, Springer Science+Business Media, May 2021, https://doi.org/10.1186/s13635-021-00117-2.
Shullani, D. et al. (2017) “Vision: A video and image dataset for source identification,” EURASIP Journal on Information Security, 2017(1). Available at: https://doi.org/10.1186/s13635-017-0067-2.
Thakar, A.A., et al. “An Empirical Study Illustrating Effects on Hash Value Changes in Forensic Evidence Appreciation,” International Journal of Science and Research (IJSR), 10(4), 2021, pp. 1356–1357. Available at: https://doi.org/10.21275/SR21501151311.
How to cite this article?
| APA Style | |
| Chicago Style | |
| MLA Style | |
| DOI | |
| URL |