Martino, some years have passed since we last talked with you! What’s changed for Amped and its customers? What’s the same?
Yes, time definitely flies. In the past few years we have grown a lot, as of today we sold our tools in 100 different countries and our team is now 30 people, but our focus on image and video forensics, and the commitment to the best user support and experience stays the same.
Our tools have grown hugely over the years and we keep adding features and improving the performances with each release.
Amped FIVE now has more than 140 different filters and tools to face any possible challenge during investigations. We support some quite hard proprietary CCTV/DVR video formats that nobody else is able to convert, we have added extremely useful tools like Perspective Stabilization and Super Resolution, and many features on the presentation side, like Timeline, Freeze Frame, Annotate and Audio Redaction.
With Amped Authenticate we started to support features related to video authentication (in addition to the extremely wide image authentication toolset) both for tampering detection and camera identification.
But the biggest evolution is related to our product line-up. While traditionally our tools have been dedicated to forensic video analysts and experts, a few years ago we launched Amped Replay, a software for accurate video evidence playback and processing for front-line officers and investigators. As many other vendors in parallel forensic disciplines, we realized that the first steps of an investigation are the most critical, and often they are not done by the analysts but by the people on the field. We created Amped Replay to help them perform the first stage of investigations in an easy and forensically correct way.
Forensic digital image and video analysis seems to be becoming more mainstream as these types of data are now consumer-driven as much as commercial. What challenges does this trend create for customers, and what opportunities?
I think the biggest issue, by far, is the lack of education on the challenges of image and video analysis. Nobody would think to perform a computer or mobile forensics analysis (and even less a DNA or ballistic analysis) without giving people the right tools and the relative training. But this is what happens routinely with video evidence. Just because somebody is able to play a video or look at a photo, it doesn’t mean they are also able to properly analyze it. Often the need for the proper methodology is overlooked in our field, giving rise to catastrophic results. We have dedicated a full series on our blog to this topic, titled “Video Evidence Pitfalls: because you don’t know what you don’t know” and we are hosting almost every month a free webinar called “Getting Started with Video Evidence” to educate the industry at large.
In 2014 you told us, “People expect to have an automatic software that will tell you which part of the image has been modified, with 100% reliability, for any kind of low resolution, highly compressed pictures downloaded from the web.” As “deep fakes” have become part of social discourse, do you still encounter those expectations, or have they changed?
Of course, people always expect the single-magic-button-software (we even had an April fools joke on this many years ago), and while we improved the automation in several aspects of our tools (for example with the Smart Report in Amped Authenticate, and the automatic correction tools in Amped Replay), the user interpretation and the scientific workflow are still fundamental to spot possible issues and reach the final result. The field is too complex, and the stakes are too high to leave the last word to an automated system. We have historically been very much against artificial intelligence (AI) in forensics, but over the years we have done research on it with a lot of experimentation; we now believe that AI can be applied to some contexts , but within very well defined boundaries and with as much understanding of what’s happening as possible. Deepfakes detection is one of these applications. Being created with AI techniques, the natural (pun intended) way of detecting them is based on AI too. A battle of two AIs sounds pretty epic, isn’t it?
What do customers actually need when they encounter digital multimedia for analysis?
This year at our Amped User Days we asked our users to fill a survey and we got very interesting insights. These are the most common activities sorted by frequency:
- Image or video enhancement
- Play or convert proprietary CCTV/DVR video files
- Create demonstrative videos or comparison charts
- Perform comparative analysis (clothes, faces, cars…)
- Video redaction
- Image or video authentication
- Taking measurements (such as a subject’s height) from images
- Measuring vehicle speeds from videos
Images and videos are almost never of a satisfying quality and if you are working on CCTV you need to convert the format to a playable one. These two tasks are almost always needed, then all the other tasks are more specific and depend on a case by case basis.
It’s a little more than a year since the COVID-19 pandemic sent us all into lockdown. In hindsight, what were the biggest impacts to customers and their needs? How was Amped able to respond?
By far the biggest issue has been operational: our users (in many cases) had to work from home as much as possible and were not able to travel. We were able to respond to these changes quickly, in several different ways.
Traditionally, our software was dongle based, and because of this it was complicated for several customers to take it out of the office, so we transitioned to digital licensing as default, removing the need for the hardware component. This also helped for shipping, in the worst part of the lockdown it was an issue even to deal with couriers in many parts of the world. We also offered existing customers free temporary work-from-home digital licenses, to help them cope immediately with the situation and allow them to work from home with ease during the emergency.
Another initiative was the launch of online training. As for digital licensing, it was something we wanted to do for a long time, but we didn’t have the time to do it. The emergency situation moved this, together with many other things. At the beginning we felt some resistance, people were hesitant and waiting to be back on the road in a few months. When they realized it wasn’t going to be so fast, online training really started to fly. Ultimately last year we trained more people than any other year, and this one we’ll train even more.
Another thing we have been missing were in person events, and while we are looking forward to be back at shows (and are already for some, especially in the US), we started doing a lot of webinars, which were way more successful than expected.
Over the past 7 years, what stands out to you? What are your biggest lessons learned, and what are you proudest of?
I think the most important thing we got is a better understanding of us as a company and as a team. We are still having the same values and the same overall strategy, but we were able to lay it down more explicitly rather than pass it over to new employees by context only. This is very important as the company grows, especially with fully remote workers. Some concepts and values may be obvious to someone who has been here for a while, but for newcomers having them clearly outlined makes things much easier and avoids misunderstandings. As a company we have a pretty particular culture. Internally the most important point is the aim to create an enjoyable workplace and being enthusiastic about what we do every single day. Externally the priority is to be fully aligned with our customers and provide them the best possible service.
To be faithful to these principles I’ve said no to several partnerships and deals, but I think this was the best decision for both us and our users, even if “on paper” it would have been a great opportunity.
What’s next for Amped? What are you most excited about for the future?
If you have been following us for a while, I think the trajectory is set. We all love our work and our customers are very happy about what we do. I don’t see any revolution, just an evolution. We have so many great ideas for our products and training that are hard to prioritize. We are exploring some new tools, but still in our expertise area. I think for a company like ours the focus is extremely important. We are doing one very specific thing and we want to keep being the best in the world at doing it, always guided by our North Star “justice through science”.