Computer vision: today and tomorrow
by Clemens Niekler, on Nov 15, 2019 8:47:00 AM
During our last Hackathon, we experimented with Google Cloud Vision API and its possibilities. It turned out that it is already possible to detect certain ingredients on the pizza and classify them by simply analysing pictures. Of course, this is not the only way this technology can be used. Google Vision AI and computer vision, in general, allow for many new application areas and will most probably change the way we will see the world in the future.
What is computer vision?
Computer vision aims for computers to understand the information given by visuals like photographs or videos. As this skill is essential even for children, it sounds like it should be easy to teach it to a computer as well. Nevertheless, many difficulties arise.
The visual world is not only black and white, nor it contains only 0’s and 1’s. It is much more diverse. Pictures are taken from all kinds of angles, under different lighting conditions, with (partially) hidden objects. Only artificial intelligence and in particular, Machine Learning can teach and train the computer on the millions of different images.
Use cases of computer vision
Computer Vision tools like Google’s Vision API and AutoML Vision are already being used in big corporations like The New York Times. The vast amount of pictures of The Times’ photo archive is stored in a room close to the Times Square in New York City called “The morgue”. All of these pictures that represent more than 100 years of recent history were digitised with Google Cloud’s technology. After doing so, Google’s Vision API and AutoML Vision were used to categorise the photos and make them easily accessible.
Environmental changes tracking
Furthermore, Google’s AutoML Vision is already used to protect the environment. The Texas A&M University uses the technology to indicate the so-called Environmental Sensitivity Index (ESI). The ESI helps to protect the sea and its marine life in case of possible oil spills. The model trains itself and thus improves at the detection of changes in the coastline.
Another example of an already existing use case of computer vision is the protection of wildlife around the world. The Zoological Society of London (ZSL) uses AutoML Vision for their camera traps that record thousands and millions of pictures in conversation areas to collect data. Many of these pictures are taken even though there are no animals around. Going through all of these photos took months. Now it is possible to do so within days, which allows the environmentalists to focus on the protection of endangered species.
From more accessible storage of recent history to environmental protection, computer vision and in particular, Google’s AutoML Vision and Vision API are used already. Not to forget are everyday use cases of computer vision. For example, many smartphone users use FaceID multiple times a day to unlock their phone, utilizing computer vision.
The future of computer vision
According to research, the computer vision market will reach $24 billion by 2026, which is twice as much as in 2017. The use cases of the technology will get broader and will most probably be implemented in almost every part of our lives at some point. Computer vision models won’t only be more uncomplicated to train in the future, but they will also be able to distinguish between more details.
One of the most prominent use cases of the computer vision technology in the future is the healthcare sector. There are already models that can detect cancer tumours more accurately than human doctors. Diagnoses executed by machines allow the specialists to focus on human contact and spend more time with their patients.
The growing field of autonomous cars is another field where computer vision plays a vital role. In the long run, it is imaginable that the technology will allow for superintelligent AI that sees the world as well or even better than humans.
Overall, computer vision might change the way we see the world and will allow humans to focus more on important tasks at their workplace and in particular, human relations.
Want to read more?
Crystalloids helps companies improve their customer experiences and build marketing technology. Founded in 2006 in the Netherlands, Crystalloids builds crystal-clear solutions that turn customer data into information and knowledge into wisdom. As a leading Google Cloud Partner, Crystalloids combines experience in software development, data science, and marketing making them one of a kind IT company. Using the Agile approach Crystalloids ensures that use cases show immediate value to their clients and make their job focus more on decision making and less on programming.