Skip to main content

Rev Case Study: How Axon is using Speech Recognition Technology to Improve Law Enforcement Efficiency

Administrative tasks are a necessary — if unfortunate — part of any law enforcement professional's day-to-day. But they don't have to be a time-suck. Axon Enterprise has developed all kinds of solutions to help officers become more efficient, from mobile apps for evidence capture to automatic redaction assistants to innovative report-writing solutions. The next step in their mission to improve efficiencies is leveraging Speech-to-text services powered by Automatic Speech Recognition (ASR). This cutting-edge technology can save officers valuable time in their days and let them get back to the communities they serve. 

While ASR has been around for more than 60 years, the last decade has seen rapid improvements in accuracy and uses across multiple industries. Rev, the world's leading speech-to-text service, compiled a research report earlier this year in which 79% of respondents said time savings were a huge benefit of speech-to-text services. Furthermore, 63% ranked it the top benefit. 

Creating Efficiencies with Speech Recognition

Leveraging Rev's ASR technology, Axon has launched Axon Auto-Transcribe to transform the way enforcement agencies review and transcribe evidence. Powered by AI, the solution converts audio and video footage into a time-synced hypermedia file empowering users to rapidly find information. "Axon Auto-Transcribe allows agencies to get the maximum value out of their audio and video evidence," says Noah Spitzer-Williams, Principal Product Manager, Auto-Transcribe. "It transforms the way agencies review evidence, allowing them to find and review the important moments in a fraction of the time."

ASR technology is invaluable for digital evidence management and transcription. Axon Auto-Transcribe's hypermedia experience can speed up the time it takes for officers to review audio or video evidence by generating a scannable and searchable transcript time-synced with the digital evidence. This enables users to quickly find critical moments within audio or video files during an investigation. 

By starting with an auto-transcript, we give users in an agency a "head-start" to create transcripts quickly and accurately. In addition to speeding up the time it takes to generate a court-ready transcript, users can avoid watching potentially harsh or disturbing video footage over and over again. 

Plus, an integration with a speech recognition leader like Rev saves officers the tedious task of logging into yet another system. "Most officers have to juggle as many as five systems and logins," says Noah Spitzer-Williams. "By having a single platform, single log-on, single workflow for evidence reviews and transcription - agencies are both saving time and increasing security."  

What Lies Ahead?                    

Axon Auto-Transcribe technology can also significantly streamline report writing. With ASR-integrated technology solutions, officers start with an auto-transcript of the incident, to ensure they provide an accurate account of events in the report. This gives officers hours back in their days that they can spend serving the community. 

Of course, none of these time-saving benefits matter if the ASR output is inaccurate. When it comes to good police work, accuracy is paramount. Rev boasts the most accurate automatic speech recognition engine on the market today, besting similar technologies from giants like Google, Amazon, and Microsoft.

An ASR engine's accuracy improves with the more data sets it processes, and that means big things for the future of the industry. Statistics show that, from 2020 to 2025, the speech and voice recognition market will leap to an estimated value of around $31.82 billion. 

And while the industry grows and the technology advances, customers' expectations of ASR are shifting. Consumers want accurate speech-to-text conversions, even if the audio's quality is lacking. Developers have tested on audio from a long distance, or in low-quality conditions (significant background noise, etc.). 

Furthermore, Ivy-League institutions like Cornell University are researching developments in speaker diarization, or the ASR system's ability to identify who is speaking in multi-speaker audio and properly attribute the text. These capabilities could prove extremely useful for law enforcement officers, further streamlining the process of reviewing interviews, evidence, and much more.

Axon is a trademark of Axon Enterprise, Inc., registered in the United States and other countries.