Intel to Streamline Computer Vision With Its OpenVINO Toolkit
Posted by Kayla Matthews
Many people primarily think of Intel as a computer chip manufacturer, but it’s on the cutting edge of technology in other ways for its role in building PCs, software, 5G connectivity solutions and much more.
Recently, Intel announced plans for its OpenVINO (Open Visual Inference & Neural Network Optimization) toolkit, which seeks to promote faster development for computer vision applications associated with edge computing.
Edge Computing Versus Traditional Cloud-Based Data Processing
Not long ago, people were solely familiar with traditional cloud computing, which requires sending data to distant locations for processing. However, there are numerous issues with that approach, including those associated with latency.
The explosion in popularity of new smartphone technology and innovative IoT devices means more data than ever gets stored in the cloud, and there are no signs of the amount reducing anytime soon. Edge computing is a method involving data processing that happens near the network where it gets taken in, thereby reducing the data center load and leading to applications that are more responsive due to fewer delays.
What’s Vision Computing?
Even though you may not have known the technologically known name for it, you’ve almost certainly encountered vision computing — also referred to as computer vision — in everyday life.
For example, newer security system apps allow you to access camera footage from your smartphone, which is due to vision computing. Also, many banks allow customers to remotely deposit checks by taking pictures of them for verification purposes. Vision computing makes that possible, too.
Artificial intelligence makes computers do some things better than humans. Image analysis is one of them, and that’s because of computer vision.
Vision Computing for Processed Video
The goal is for Intel’s OpenVINO toolkit to specifically target video processing and make it more efficient. The previously mentioned latency and data load issues mentioned with cloud computing make it challenging to scale up and analyze more data points within videos.
Also, edge computing lets companies retrieve high-quality and uncompressed video streams, allowing representatives to hone in on the details that matter most to business operations.
Intel’s OpenVINO toolkit made the news and earned attention from IoT enthusiasts who realized the upcoming tools would let developers build image-recognition models compatible with numerous Intel chips. Using the OpenVINO toolkit facilitates adapting AI models for various uses and devices with less time consumption and effort.
Intel's vision is not a one-size-fits-all approach, but instead hardware and software that adapts to the situations that need it. OpenVINO fits into that aspiration and will be particularly advantageous for business and security insights.
The OpenVINO toolkit is among Intel’s offerings that allow identifying faces, traffic patterns and even license plates in video footage. Then, businesses and security experts can take that data to determine things such as the busiest times of the day — or the gender of people most likely to shop at a particular establishment on Monday afternoons, for example. They could also use it to gain intelligence about suspected shoplifters or vandals.
