Computer Vision: The Rising Tech Of The Post Corona World
The surge in the number of COVID-19 cases globally has changed certain practices that we’d taken for granted, especially involving touch-based sensor recognition, aka biometrics. From workplaces that have gone remote, to manufacturing units adopting robots, to mechanized services disrupting the supply chain logistics, entire processes, including last-mile delivery, have undergone massive changes in an extremely short span.
In this context, all fingerprint-based biometrics have been suspended, and enterprises have moved to facial recognition, especially for identifying employee counts and maintaining security access! Why even New York’s premier law enforcement agency NYPD has also called for the stalling of fingerprint-based scanning systems. The time is now ripe for contact-less, computer vision tech to take over.
Computer Vision has Extensive Scope
CV is the science and technology that allows machines to see. It’s interesting to note that we may still be in the nascent stage of growth, given that the market for computer vision is projected to reach $26.2 billion by 2025! Stanford University’s leading AI Scientist Fei-Fei Li feels that vision is the most competitive cognitive ability that can change the course of AI. She goes on to add that ‘Understanding vision and building visual systems are really about understanding intelligence.’ This is undoubtedly true as major tech giants of the world – Google, Facebook, and Microsoft are heavily invested in computer vision technology in various capacities to make it real.
Typically, computer vision technology uses special neural networks called Convolutional Neural Networks, which are trained using thousands of sample images and algorithms to break down individual components in the image pixel by pixel, and these are then memorized by the system. This becomes the database and reference, and with every cycle of iteration, the AI system learns on its own and becomes proficient! Here’s one of the most common infographic depicting the basics of computer vision:
Widespread Applications of Computer Vision
It wouldn’t be an exaggeration to say that we’re living in times of a data deluge – the ubiquitous smartphone within our homes often complemented by equally smart televisions, voice assistant technologies, CCTV cameras in the streets, public spaces and offices, and law enforcement agencies. There’s an explosion of sound, texts, images, and videos. A report by “Facial Recognition Market” by Component, pegs the industry growth from $3.2 billion in 2019 to $7.0 billion by 2024 in the US. To be fair, Computer Vision as the software has been at a tipping point and will reach its peak due to the hype cycle in the next few months, undoubtedly spiking massive advances in deep learning algorithms and graphic processors.
On an individual level, you only have to look at your smartphone galleries to know how many apps could be leveraging it! This also points to increasing acceptance and voluntary participation from individuals despite privacy issues and regulations. There is a growing comfort among individuals to experience a more efficient way of working, and hence adoption is surmounting rapidly.
COVID-19 has ushered in the era of digital transformation like no other and has spurred enterprises to adapt to the new age of working to remain relevant. This major shift in the way of working has forced enterprises to re-look at core processes while innovating on newer business models.
Computer Vision software is already being used by several industries across multiple domains spanning financial institutions, consumer packaged goods, insurance, etc.
- Healthcare: CV tech is being used increasingly in early-stage disease detection and diagnosis. In general, the emerging global health crisis has necessitated a change in how even limited human interactions are evolving. Doctors are using AI Models that deploy deep learning methods on image classification in examining hundreds of CT Scans to speed up the diagnosis by distinguishing Coronavirus affected tissues from pneumonia cases. Now, temperature-detection algorithms are being bundled with facial recognition software to check two things off the list: Attendance in enclosed spaces as well as people’s safety. Given that masks may become an integral part of the future even after Corona abates, experts will have mastered facial recognition algorithms that work even when respondent faces are partially blocked – making safety paramount!
- Manufacturing: Computer vision in manufacturing is paramount to a more efficient process. Whether it’s in inspecting packages, scanning barcodes, tracking worker safety, etc., there’s a lot that can be done to enhance existing workflows. With COVID impacting conventional manufacturing processes, the few medical companies who are still producing essentials such as Personal Protective Equipment, Pharmaceuticals, or Sanitation Products can supplement their workstreams with CV algorithms that combine RGB and thermal cameras to identify people with high temperature. Even QA teams across factories can adopt CV technology for the final stage inspection to ensure efficiency and avoid human interactions.
- Automotive Industry: One of the most promising use cases of complex CV algorithms is in the research and development of driverless vehicles which use maps of the surrounding areas, various sensors and object recognition software for movement. Despite extensive research by Tesla and Waymo, safety has been a major concern that’s halted the adoption of these vehicles. However, a Beijing based self-driving logistics start-up – Neolix Technologies has confirmed funding of $29 million this February towards mass production of its self-driving cars. Delivery Robots are seeing a sudden surge due to the coronavirus crisis. Kiwibot is one such robot that’s been delivering medical requirements such as masks, antibacterial gels, and hygiene products for the communities of Berkeley and Denver.
- Retail: The Coronavirus crisis is one of the worst catastrophes for the retail industry. In the times to come, Computer Vision technology can be a real game-changer among brick and mortar stores. Amazon Go’s ‘walkout technology’ uses CV algorithms to understand which items have been picked up by consumers and auto deducts the amount from their wallet – making it an excellent example for contactless payment. In general, the retail giants can benefit from using facial recognition to identify and segment consumers based on purchase patterns and personalize their shopping experience.
- Banking and Insurance: Opening accounts at financial institutions often takes an unnecessarily long time. Enter facial recognition wherein customers can directly fill forms online and update their images by uploading selfies. The details are verified within a minimum timeframe, and the account gets authenticated. While Coronavirus hasn’t impacted the BFSI as directly as other industries, there’s still scope for new-age technologies to revolutionize this space. Conventional practice demands that in insurance underwriting, claims are filed, the surveyor visits the scene of accident/workshop to assess the damage, the claim is evaluated, and then finally, it may be approved. The entire process easily stretching into weeks, if not months and this can be done for personal vehicles involved in accidents or to property.
BRIDGEi2i partnered with a global financial institution and initiated drone-based image capture to assess the damage done to rooftops, thereby reducing the time for actions and managing to achieve 95% accuracy for the same.
BRIDGEi2i’s robust AI & Innovations Lab team has a dedicated division of experts who work on multiple projects as well as research. We partner with several Fortune-500 companies and are striving hard to do the best we can in the pressing times of COVID-19. Our AI-powered accelerators are plug and play solutions that work on advanced AI algorithms to ensure faster time to value and improved decision-making across the board to make AI Real for Enterprises!
In this regard, we’ve been analyzing the global pandemic and developed the COVID-19 Global Monitor, which collates information across news sources in a comprehensive manner. This, in addition to the BRIDGEi2i Market Impact Monitor, is among our foremost contributions to the fight against Corona. Take a look below!
Author: Spurti Devadhar
Tune in here to listen to “How Data & Analytics Can Help in a Downturn”