Despite its recent surge in societal familiarity, Artificial Intelligence is actually something that has been around for over half a century now. It was first explored as early as the 1950s, which was a decade that saw two pivotal moments in the realm of AI discovery. Alan Turing, the computer scientist who famously cracked Germany’s ‘Enigma’ code and played a pivotal role against the Axis Powers, was arguably the pioneer of the AI that we know, love and fear today. In 1950, Turing posed the tide-turning question ‘Can machines think?’ in a study known as ‘Computing Machinery and Intelligence’. It was here that he invented the self-described ‘Turing Test’, an experiment of sorts that could determine whether a machine was capable of thinking in a way that could imitate the thought process of a human being. At the time however, no one was referring to it as ‘artificial intelligence’. That infamous term was only coined a few years later by Professor John McCarthy in 1956. Since then, the neural networks of machines have only expanded their influence into our everyday lives. One such example is the realm of photography.
AI’s involvement in photography is nothing new, though for the most part its role in the industry was historically limited to just post-production tasks. Creative programmes like Adobe Photoshop and Lightroom have allowed photographers to experiment with editing tools such as blur and noise reduction, image sharpening and digital enhancing. Today however, the progression of AI has accelerated into another crucial part of photography – the camera itself. AI has taken on various duties inside the camera, many of which were once delegated to the photographer themselves.
Multiple tasks such as motion tracking, face and eye detection can be handled entirely by AI, without any input from the photographer whatsoever. Most notably, the advancements in AI have allowed auto-focus to reach a higher level of accuracy than ever before. In fact, the success of the mirrorless camera can in many ways be attributed to these new features. Every high-end full-frame mirrorless camera has adopted AI in some way or another, proving that the technology is not only here to stay but is integral in overall photographic production now.
For example, Sony’s A7R V camera made waves in the industry when it was first released back in 2022 as it was the first camera in their alpha range to make use of deep learning AI subject recognition, providing photographers with fancy new tech that delivered real-time subject tracking as well as other enhancements like wider auto-focus coverage with more detection points, better auto white Balance and low light performance improvements. As such, the A7R V was highly commended for these features, yet interestingly Sony was not the first major brand to implement this. Technically speaking, it was Canon who got the ball rolling when they released the EOS-1D X Mark III all the way back in January 2020, a year many of us will be cautiously familiar with.
The Mark III was Canon’s first foray into the world of deep learning algorithms, signifying a shift in industry views towards AI. In more recent years, Canon’s EOS R1 and EOS R5 Mark II have continued to demonstrate the brand’s commitment to these algorithms, with both devices featuring new tools such as the ‘People Priority’ mode and ‘Dual Pixel Intelligent AF’ which allows the sensor to continue tracking subjects at higher speeds, even when briefly concealed behind obstacles. The People Priority mode on the other hand, is exactly what it sounds like, with the faces of the most prominent people in the image reciving a boost in clarity and overall sharpness, making it an invaluable feature for photographers at large-scale events.
One area that continues to grow is the smartphone camera – not just in its use of similar facial-recognition features, but in sheer pixel quality. Mirrorless cameras might offer a superior image density, but that doesn’t change the fact that the iPhone 15 Pro Max for example, boasts a 48-megapixel sensor, practically rivalling former professional cameras like the EOS 5DS. Such key technology poses an inevitable dilemma however – does AI completely remove the need for creative talent, and negatively impact the photographer’s skill? While it is true that the basis of AI is to perform tasks that would otherwise require human input, it does indeed deduct many elements that photographers have historically stifled over – manual focusing, image framing and depth of field are all vital to communicating a photographer’s style, but if these duties are solely carried out by machines then a lack of engagement on the photographer’s end would surely be expected.
The manual focus pull of the lens is one of the most satisfying experiences a camera can offer, but even then, a photographer could be missing other elements in the frame. Facial features like bone structure and eye sharpness can be automatically highlighted in certain cameras, with the EOS R6 Mark II even introducing the option to specify left or right eye priority. Other AI-led detection features are also present in the EOS R10 and R7, which give the photographer the choice to quickly select one person from several and lock the AF tracking frame on that specific person. It is clear then, that auto-focus does offer up speed and convenience, but for many, manual focus still provides the ultimate precision which is particularly important in certain low-light environments.
In conclusion, AI in cameras is much more than a trendy gimmick, it is reshaping the very images we produce – editing tools like generative fill can seamlessly fill areas of an image after removing unnecessary elements, achieved by analysing the colours, tones and lighting to artificially guess what’s behind the object. It is of course no secret that AI played its role in the death of the DSLR, and for many that alone is a point of contention. But to resist would be futile. After all, AI will continue to learn whether or not we learn with it, so why not embrace it?
Source link