AI And Machine Learning: The Future Is Now
It seems that in the last few years Artificial Intelligence (AI) has really come of age. With machine learning at the heart of the latest advances in AI, researchers have acquired new knowledge and methods to achieve impressive results across a wide range of areas including image and video recognition and analysis. And what’s more, these techniques and results are now within easy reach for investigators and their teams.
By Johann Hoffmann, Head of Griffeye, and Pelle Gara, Director & Co-founder of Griffeye.
Over the last couple of decades, AI has always been a hot topic however the potential has often been more exciting than the reality. But things have changed over the past four or five years. From Siri on our iPhones to increasingly sophisticated Google searches and Netflix recommendations based on what we’ve already enjoyed, AI and machine learning are being brought into our everyday lives to provide quicker and more accurate assistance to people. No longer restricted to the research lab, we’re now seeing the democratization of machine learning through easier access to the techniques and an explosion of applications.
Deeply intelligent and reliable
Within the field of image recognition and analysis, recent advances have been made thanks to deep learning techniques and the development of Convolutional Neural Networks (CNN). These are inspired by how our own brains work so efficiently to recognize objects – by filtering the lines, then the shapes and finally the objects themselves in a hierarchical approach. Of course the advantage for computers carrying out human tasks is that they can process many more images and quicker, plus they are as “focused” on the 372,487th image just as much as on the first. And with video analysis there’s incredible potential when you think of the ability to quickly go through thousands of hours of material and still pick up the tiniest clues, something that’s almost inconceivable even if you used teams of people working around the clock
The democratization of machine learning and the newfound access to these techniques is well illustrated by a recent story. Washington County Sheriff’s Office in Oregon, USA needed a quicker and more accurate way to identify suspects from images. This used to involve sending e-mails to law enforcement officers to ask if they recognized the person or people in the image. Clearly, this could result in delays before officers were on duty and answered the e-mail or perhaps they just wouldn’t remember.
From months to minutes
But then Chris Adzima, a senior information systems analyst at the Sheriff’s Office, heard about Amazon’s recently launched Rekognition, a deep learning-based image and video analysis service and wondered if that could be worth a try to help them identify suspects. He uploaded all the mugshots they’d archived over the past two decades, and using that as a reference catalogue quickly set up a powerful tool using the Rekognition API for identifying suspects. Now, instead of days, weeks and sometimes months waiting for answers to their e-mails, the Sheriff’s Office is getting positive identifications within minutes – they’re even getting results from artist sketches!
With the time and resource pressure on investigators and their teams, AI and machine learning can help massively. Now that they’ve become more democratic through easily available APIs (supported by Griffeye’s own open API) and training resources, it’s simple to get started and reap the rewards.
Welcome to the future!
Further reading and resources:
— Read the blog or watch the talk about Washington County Sheriff’s Office use of Amazon Rekognition.
— Visit Microsoft Cognitive Services, a similar service in the democratization of machine learning.
— To learn more about how to build your own classifiers with the help of Deep Learning, we recommend www.fast.ai, an excellent and free online course.