Author Archives: Marco Fontani

Amped’s New Deepfake Detection Algorithm Published in Journal of Imaging

Dear Amped friends, we are glad to share some good news with you all. After months of complex research and testing work, we developed an effective deepfake detection algorithm. The technique can detect GAN-generated images, that are, synthesized faces “invented” from scratch by a suitably trained neural network (check this article for an introduction to deepfakes: Dealing with Deepfakes). Currently, it has been tested on images only.

Continue reading

Amped Authenticate 26549: New Face GAN Deepfake Detection Filter, Updated Social Media Identification Database, and More

November comes with a great Amped Authenticate update! We are adding deepfake detection capabilities, which increases the already broad coverage of Authenticate for localizing forgeries. Moreover, we’ve updated the Social Media Identification‘s signature database with hundreds of new signatures and added one more social media, or rather a messaging app: Facebook Messenger. In this blog post, we’ll first quickly review what deepfakes are, focusing on synthesized faces, and then we’ll go practical and discuss each improvement!

Continue reading

Does Deep Learning-Based Super-Resolution Help Humans With Face Recognition?

Dear Amped friends, we’re glad to announce that our research team, in collaboration with prof. Sergio Carrato (University of Trieste), authored a new scientific article! The title is the same as this post’s: “Does Deep Learning-Based Super-Resolution Help Humans With Face Recognition?”.

There is little doubt that AI-based super-resolution images (bottom left) are visually more appealing than bicubic-interpolated images (bottom right). But does that lead to improved recognition accuracy?
Continue reading

The Full Workflow of Speed Estimation 2d in Amped FIVE

We hope you’re aware of the big news: a few weeks ago, an extraordinary Amped FIVE update rolled out bringing Speed Estimation 2d filter into the software. We’ve dedicated a blog post to explain what the filter does and how to use it, and we recommend that you read that post before going on with this one. Hereafter, indeed, we’ll be explaining how the filter works, how we compute the speed and the uncertainty. Happy reading!

Continue reading

Authenticate Update 23481: Improved Variation of Prediction Footprint Analysis and More

As we hope you know already, at the end of 2021 we’ve added the Variation of Prediction Footprint (VPF) video tool to Amped Authenticate. VPF allows detecting video double encoding (if you missed that, have a read here!), which is massively important for video integrity analysis. This early update adds more features to VPF analysis and improves the interaction to make your workflow easier and quicker!

Continue reading

How To Check Video Integrity By Detecting Double Encoding With VPF Analysis

This article will start by defining video integrity and mentioning possible approaches to integrity verification. We’ll also make a very short introduction to some concepts of video coding and then, show how you can use Amped Authenticate’s VPF analysis to reveal traces of double compression in videos, and how important this could be as an integrity verification step.

When dealing with video evidence, integrity verification is a fundamental, yet often neglected, task. The definition of integrity can be summarized as: “the file is original and unaltered since the time of acquisition”. It’s plain and simple: the file should not undergo any change once it’s acquired, and if it does, every change should be documented in the chain of custody. Notice that integrity is different than authenticity, which has a more “semantical” meaning, as it requires the video to be an accurate representation of what it purports to be. If you re-compress a video, its authenticity will normally be preserved, but integrity is broken.

I know someone could think “the last thing I need is having to waste time checking one more thing in all my casework”. Understandable. At the same time, though, if one day I’ll be charged with something, I would really love to know that all the evidence pertaining to my case has been handled with the greatest care and I’m quite sure that holds for all readers of this article! If you’re in the forensic field, then you know that working with video is not just “common sense”. We’re not sorting our holiday pictures; we are handling evidence that could change somebody’s life, and dramatically so. 

Let’s take a deep breath and dive in!

Read the full article here.

Amped Funds a 3-Years Research Grant at the University of Florence

Since Amped lives on technology, it’s evident that research is essential to our company. “Research is important” is too often just a mantra or something companies and institutions like to repeat. But if you genuinely believe in that, the only way to go is to put resources on it!

We’re thus thrilled to announce that Amped Software partnered with the Department of Information Engineering of the University of Florence (Italy), and funded a 3-years Research Grant! A candidate has been selected and has already started working under the supervision of professor Alessandro Piva, a world-renowned expert in the field of multimedia forensics.

The research project covers video forgery detection and integrity verification, forensic analysis of videos encoded with state-of-the-art algorithms, and more.

It is a significant commitment, and we’re very proud of fostering academic research in times where the sanitary crisis has raised budget concerns virtually everywhere in the world.

Stay tuned; we’re sure exciting results won’t be long in coming! You can also follow us on LinkedIn, YouTube, Twitter, and Facebook so you’ll be promptly notified of our next achievements!

Amped Authenticate Update 22874: Introducing VPF Analysis for Video Double Encoding Detection and Improvements to the Shadows Filter

Have you ever heard “Dulcis in fundo“? It’s a Latin saying that means: “leave the sweetest for the end”. And this is what we’ve done with Authenticate this year! Just before this challenging 2021 reaches its end, we’re sharing a new Amped Authenticate update with an important new tool. Keep reading!

Continue reading

More Real Than Real: a Chat About Deepfakes with Dr. Cecilia Pasquini (University of Trento)

Deepfakes and computer-generated images have been around for a few years now, and they’ve become quite popular. We’ve recently authored an article in the Evidence Technology Magazine about the challenging task of Dealing with deepfakes. Much like it happens for cryptography-vs-cryptanalysis, a fight is going on here between those developing advanced neural networks, capable of creating more and more realistic fakes, and those aiming at detecting them. This fight is mostly guided by researchers, but video forensic analysts and everyone dealing with forensic image analysis are necessarily involved, since they may soon have to face deepfakes in their investigations!

Continue reading