Interview with Marco Fontani, Forensics Consultant, FORLAB

This is the first in a series of interviews with a number of our users that will share their story. If you are interested in being profiled, let us know. We would love to hear from you!

In this first post, we speak with Marco Fontani from FORLAB, one of the main laboratories in Italy specialized in image and video forensics. We started working together in the European project MAVEN and since then, we cooperate with FORLAB on several fronts, including the technology transfer from the world of research into practical applications.

Marco, tell us a bit about yourself. What is your background and your current role at FORLAB?

My background is mostly in academics. I have a Master’s degree in Computer Engineering and a PhD degree in Information Engineering and Sciences. I’ve authored and co-authored several publications on novel image, video and audio forensic algorithms. My current role at FORLAB includes a mix of consultancy and training plus some research activity (mostly carried out by following master/PhD students).

What made you decide to enter the field of multimedia forensics?

I’ve always been interested in multimedia security because multimedia content is much too easily trusted/relied upon. When I began my PhD in 2010, an important research project funded by the European Commission (REWIND project) was about to start, and I joined the research group of professor Barni to take part in this project. After the PhD, I thought it would be interesting to put my studies to work in practical cases. Hence I, together with other colleagues, pushed the growth of FORLAB, where the scientific and technological innovative results are transferred to the real needs of the forensic environment.

What would you say are the biggest challenges with multimedia digital evidence and investigating crimes?

In the age of terrorism, I believe the main challenges are related to: a) analysis and interpretation of massive data (e.g. images/videos shared through social networks, but also video surveillance footage captured by cameras spread in a city); b) poor quality of CCTV cameras, that makes recording useless in so many cases.

What would you say are the main forensic challenges surrounding image validation? How can they be addressed?

Some challenges I see:
a) Creating algorithms whose output is easily interpretable (and reliable).
b) Designing a comprehensive analysis workflow that helps the analyst make sensible choices based on progressive interpretation of results.
c) A huge challenge is related to images whose source and processing chain is not certain (e.g. all images coming from the web/social networks). I think there is no easy way to tackle this problem. A good starting point is to be able to recognize at least the last step in the image history (e.g. understanding that an image comes from Facebook/Twitter/etc.).

In your opinion, how important is it that digital forensic techniques and tools are based on the scientific method?

I believe this is of paramount importance.

Why do you think that it is so important that forensic image and video analysts be properly trained?

I think training is important because processing/analysis tools may be dangerous if you don’t know the basics of image and video formation and coding: it’s very easy to misinterpret findings and reach the wrong conclusion. Training also allows you to remain up to date (if renew from time to time), and can also help save time (e.g., because you are taught how to quickly understand whether something can be done or not).

How do you think the world of image and video forensics will change over the next few years?

Standard cameras are being replaced by smartphones and tablets, and producers invest a lot of resources to make higher quality pictures from a camera hardware that is not comparable to that of DSLR cameras. This means that “strange tricks” are used both at the hardware and software level, which makes the acquisition chain more and more complex. Creating proper models for such complex processing will be a real challenge.

Deep learning will also raise many questions: lots of novel scientific contributions are based on deep learning today, but training-based systems are often questionable in practical forensic cases.

Video forensics has even more challenges: 4k videos are entering the game, while scientific literature still makes experiments on CIF resolution videos because algorithms are computationally heavy. Moreover, the continuous development of coding standards makes it harder to find long lasting reliable methods (e.g., many methods that work for H.263/H.264 won’t work with H.265).

How did you learn about Amped Software?

I’ve worked on a research project (MAVEN) in which Amped Software was one of the partners.

Do you have any interesting story or success case related to Amped Software products?

Amped FIVE helped us in many cases. One interesting case involved a homicide where the height of the suspect was of interest: using Amped FIVE we could restore the video (removing lens distortion and some compression artifacts), enhance the video and measure the subject in several frames, yielding a reliable estimate. With license plates, we’re less lucky (but we were warned about this during training!) because in 90% of the requests we receive for license plate number enhancement there is little information available in the image. With the remaining 10%, we can normally get a partially readable plate.

As for Amped Authenticate, it proved important in several cases for its Quantization Table database and for the Camera Identification tool. We also extensively used the Correlation Plot tool in one important case in which the questioned images were allegedly down-sampled in some parts.

When you are not busy looking at digital evidence, what do you like doing in your spare time?

I like taking walks in nature, reading books, blogs, and news, and cooking. I’m not on Facebook, which is likely saving me a lot of time!


FORLAB provides consulting, training and research for the study, development, and use of cutting-edge solutions for handling multimedia contents for investigative and forensic purposes. For more information visit:

Table of Contents

Share on

Subscribe to our Blog

Related posts

Learn and solve it with Amped FIVE

Increase Exposure of Dark Footage

Happy Tuesday everyone and welcome to our fourteenth installment of the “Learn and solve it with Amped FIVE” series. This week we will discuss how

Read More »
Learn and solve it with Amped FIVE

Enhance and Optimize Facial Detail

Hello, jolly good people! Nice to see you popping back for another article of our “Learn and solve it with Amped FIVE” series. This week

Read More »