As terrorists increasingly attempt to mask their intentions through encrypted, self-destructing messenger apps and pre-Cold War era techniques, the need for improved counter-terrorism technology is increasingly expanding. One front that has become popular in recent years is the addition of voice recognition technology to take the mask off terrorists who aim to hide themselves in videos and messages.
Voice Recognition’s Success in Identifying Terrorists
Siddhartha Dhar is a muslim convert from East London who has tentatively been identified as the man behind the mask in a recent video released by ISIL. Authorities used voice recognition technology to identify the man. Voice recognition technology can be enhanced using visual tools to measure the timbre, pitch and pace of speech to better identify terrorists. By identifying key figures using voice technology, we have additional information in the search for these terrorists.
Understanding Speaker Identification
When you speak, you’re conveying several layers and levels of information that can be analyzed and dissected. The tone of voice and manner of speech is an important aspect of any conversation. The Speech Systems Technology Group at Lincoln Laboratory has created technologies that are able to not only recognize the words used, but also the language being spoken and the topic of conversation. This is crucial since technologies like this can be put to work analyzing chatter and looking for specific lingo that may provide crucial information into terrorist plots. United States experts in counter-terrorism have admitted it’s impossible to monitor all conversations, but artificial intelligence applications and advanced voice recognition technologies can help in this fight.
Speech Recognition and Human Capacity for Understanding
There are certain tasks that are very difficult for computers to accomplish. Facial and speech recognition are difficult to reproduce using computers. However, one area where computers excel is speaker-recognition. Humans have some difficulty identifying a speaker without the help of visual cues. Imagine being able to sort through thousands of terrorist conversations to focus in on key leaders within the group. Certainly, any messages conveyed by the leader of an organization will be deemed more important than other messages. This is crucial in the fight against terrorism since authorities have plenty of audio data to program computer systems to accurately detect, search out and catalog conversations by important leaders.
How Speaker Recognition Technology Works
There are several aspects of your voice that can be identified and analyzed by a computer. The unique manner of exhaling, resonation of our vocal cords, articulations, unique speaking habits and the shape of our tongue, teeth and lips are all used to create a unique signature. A computer can recognize these signatures and identify the speaker attached to the voice. Of course, it remains to be seen how sophisticated this technology gets, and there are likely ways for terrorists to change their voice to make them unrecognizable. Everything from false teeth to intentional modification of their voice can create false voice prints.
Who Is Using Speaker Recognition Technology
It’s difficult to know for certain who is using speaker recognition technology to aid in national security. However, we do know that the United Kingdom, United States and China have all used the technology on some level. While the technology is fairly new, it is being rapidly developed and improved to make it more effective in practical application. Even if the technology isn’t quite ready for large-scale pilfering of thousands of conversations, it is possible to use the technology in more direct, one-on-one conversations. For example, the technology can be used discretely at airports to identify suspicious individuals. Answer a few seemingly meaningless questions at customs, and your voice could be instantly sent to a database where a computer analyzes it and matches it with a digital voice fingerprint on file.
We’ve been using voice technology since the 1970s to aid in identifying criminals. It’s no small leap to suggest that as the technology improves, it will eventually be used to identify if an individual is present in a bar, meeting or large gathering of people. Sensors places throughout major cities can continually analyze individual voice signatures to determine whether certain terrorists are present. While we can’t always decrypt their messaging apps, it’s becoming increasingly possible to set up automated systems to detect their physical presence.