Visualri Image and Video Forensics Toolkit

In this era, digital videos and images have become a source of evidence in criminal investigations. Just like in the physical world, we leave traces behind, like fingerprints, hairs, and DNA, similarly in the digital realm, we leave digital traces such as images, videos, and metadata. These traces m

Project Title

Visualri Image and Video Forensics Toolkit

Project Area of Specialization

Artificial Intelligence

Project Summary

In this era, digital videos and images have become a source of evidence in criminal investigations. Just like in the physical world, we leave traces behind, like fingerprints, hairs, and DNA, similarly in the digital realm, we leave digital traces such as images, videos, and metadata. These traces may be helpful for legal purposes in determining the parties' activities involved in a criminal case, or even as a resource for cyber-criminals looking to reconstruct information or identify credentials on their victims. This is where Visualri, an image and video forensics toolkit, comes into play to tackle the challenges like detecting disguised absconders.

The merging of advanced monitoring security networks and ever-growing artificial intelligence leads to the emergence of an application called face detection. In this era of technological advancement, CCTV cameras are present everywhere, be it airports, offices, roads, or streets. It helps security organizations monitor the surroundings and look for suspicious people. But looking through all the available video footage is not possible. Here face monitoring comes into play. It helps in searching for the absconder.

Along with many merits, there are some downsides to every application. In the case of face detection, the system can not recognize the absconder if it is disguised or the image used for identification is not recent. Our project aims to bring innovation to face detection: use GAN to generate different disguises of the absconder and then apply face recognition to detect them. Creating fake media is very easy these days, so the credibility is very concerning. The imposter detection system will distinguish whether the media is legit or is it a deep fake. And reverse search on the images will give the sources of the image and the exif data that will help distinguishing doctored image from the real deal.

Project Objectives

The main objective is to perform Video and Images Forensics and our focus would be on the following:

• To detect absconder

• To implement deep fake detection

• Content Based Image Retrieval

Project Implementation Method

The proposed method presents a detailed deep learning-based model for face identification and detection. We are proposing this method specifically for criminal identification. An image of the person, who you want to identify is given as the input to the machine or generative network. Our goal is to determine the target’s identity by utilizing the visual clues obtained via the camera. The input image is fed into the Generative Adversarial Network (GAN) . The image is taken into the generator first after preprocessing, the image features that are visible are stored as a feature set f(x) for the input image x. The features set is used to generate labels of the same target with slightly different features. This would describe the target that they changed their appearance at present and made variations such as change of hair colour, eye colour, etc. Generative Adversarial Networks are powerful enough to identify the target even in their best disguise since their facial features remain the same. The generator module of the GAN network takes the features of an image as an input and produces a set of labels for each face found. This output image is fed into the discriminator module of the same network. It is the responsibility of the discriminator to take the generator’s output and check if the features of the given target match with any of the ones in the database. If there is a match, the image will be invalidated and the identity of the target will be returned. Thus the network will extract as many visible features as possible from the images x= 1 to n, and combine all features to produce the superset of features f(x).

For imposter detection system, deep fake detection would be applied on datasets CelebA, FFHQ, LSUN, CelebA-HQ, FoceForensics, FaceForensics++.

The CBIR will be implemented by crawling search engines like Google images, yandex, TinEye, PimEye and Bing and selecting extracting relevant info from the obtained links. This will also include extracting exif data of media will sometimes include  gps location where image was taken.

Benefits of the Project

The expected outcome of this project would be detailed analysis on the input in different domains, which would be helpful for organizations making use of it.

The main goal of digital forensics is to extract data from the electronic evidence, process it into actionable intelligence and present the findings for prosecution. All processes utilize sound forensic techniques 

It will help the forensic team to analyze, inspect, identify, and preserve the digital evidence residing in multimedia.

Technical Details of Final Deliverable

A webapp will be provided that will contain following components.

1. Absconder Detection System- Takes an image of a face and registers it in the database. Now it will detect if that person is present in the imagees that need to be searched. The face verification system will generalize in such a way that it would affectively be able to recognize the person even if he uses disguise or the appearance of the person has changed since the time the original target image was taken.

2. Imposter detection system- This basically detects if the media is real or deep fake. It catches fake and impersonated media.

3. Content Based Image Retrieval - This System get exif data of media and reverse search images on the web for passive recon.

Final Deliverable of the Project

Software System

Core Industry

Security

Other Industries

Education , IT , Media

Core Technology

Artificial Intelligence(AI)

Other Technologies

Sustainable Development Goals

Industry, Innovation and Infrastructure, Peace and Justice Strong Institutions

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com