Skip to main content

Sony wants to make cameras smarter with an onboard A.I. chip

Over the last few years, tech companies such as Google have turned to machine learning to push the limits of what a camera can do without bulky hardware. Now, Sony, that dominates more than half of the image processing market, is throwing its hat in the ring with two new sensors that come equipped with an onboard A.I. chip.

Officially called the IMX500 and IMX501, Sony says these new 12.3-megapixel sensors tackle several obstacles of how artificial intelligence is today applied to camera data. Since they have a built-in chip for the purpose, the Japan-based manufacturer claims its sensors eliminate the “need for high-performance processors or external memory” and can deliver results much quicker compared to when cameras have to transfer the metadata first to an external system.

Sony’s latest sensors run the image signals they capture natively on a separate logic chip and only beam out the extracted metadata. As the image information never itself leaves the sensor, Sony is also promising a more secure and private experience.

“These products expand the opportunities to develop A.I.-equipped cameras, enabling a diverse range of applications in the retail and industrial equipment industries and contributing to building optimal systems that link with the cloud,” the company wrote in a blog post.

However, you won’t find this technology on consumer products such as phones and handheld cameras just yet. Instead, these new sensors, for now, will be restricted for commercial purposes such as surveillance cameras and smart retail spaces that demand complex computer vision architectures such as Amazon Go.

Since Sony’s sensors can process data in real time, they won’t have to transfer every piece of footage the camera records. So for instance, at a cashier-less supermarket, the camera will only upload a particular aisle’s videos when there’s some activity instead of constantly transferring it for another A.I. computer to process — resulting in a dramatic conservation in data output and cloud servers.

Sony told Engadget that the sensors can perform “light machine learning tasks” like keeping track of the number of customers entering a store or identifying objects. Therefore, they won’t replace existing camera modules just yet. But it’s clear Sony will push ahead with more powerful and capable of A.I. sensors in the future. Sony says it has already sent sample kits to partners, so devices powered by them won’t most likely arrive before 2021.

Shubham Agarwal
Former Digital Trends Contributor
Shubham Agarwal is a freelance technology journalist from Ahmedabad, India. His work has previously appeared in Firstpost…
Samsung Home Hub uses A.I. to make your home smarter
Samsung Home Hub cooking screen.

Samsung has been connecting its appliances with the company's SmartThings ecosystem for years, and now we're seeing the culmination of those efforts with the Home Hub. Designed to help you run your home more efficiently, the Home Hub's 8.4-inch tablet brings control of all those SmartThings apps into one central location.

The Home Hub doesn't just create one place for users to monitor their Samsung appliances. It also uses A.I. to deliver a more customized solution for you and your family. SmartThings learns your preferences to make relevant suggestions. The SmartThings ecosystem works with a myriad of products, not just Samsung. Now that SmartThings also supports Matter, the number of products you can connect through the company's Home Hub just got larger.

Read more
I tested the Pixel 6 Pro’s camera against the iPhone 13 Pro to see which is best
Pixel 6 Pro and iPhone 13 Pro in hand.

The iPhone 13 Pro may be Apple’s latest smartphone, but its camera isn’t a big upgrade over the iPhone 12 Pro. So where does that leave Apple when it competes with the newly released Google Pixel 6 Pro? The iPhone and the Pixel have long fought for camera supremacy, and if the 13 Pro can’t always best its predecessor, can it at least beat Google’s latest model? To find out, I took both phones out for a couple of days to take a selection of photos, and the outcome may come as a surprise to some.
Camera specs and testing procedure
Before we look at the photos, let’s check the specs of each camera. The Google Pixel 6 Pro has a 50-megapixel main camera joined by a 48MP telephoto camera with 4x optical zoom, plus a 12MP wide-angle camera. The Apple iPhone 13 Pro has three 12MP cameras, with the telephoto providing a 3x zoom, and the wide-angle camera a 120-degree field of view.

The iPhone 13 Pro has iOS 15.1 installed, and the Google Pixel 6 Pro has Android 12. All photos were taken back-to-back, and all I did was press the shutter button. Once the shots had been taken, I compared them all on a color-calibrated monitor. You’re seeing the original photos below, just slightly resized to make them more suitable for online use.

Read more
Can Google’s Pixel 6 Pro camera beat the Samsung Galaxy S21 Ultra? I found out
Galaxy S21 Ultra and Pixel 6 Pro.

The Google Pixel 6 Pro's camera has a lot of consumer expectations behind it given the Pixel 5's success and the new Tensor processor, but it also faces a lot of work to do to compete against the Galaxy S21 Ultra, the best camera Samsung has ever put on a phone. I took both the Samsung and Google phone out for the day to take a selection of photos, to answer the question: Which Android flagship has the best camera?
Camera specs and testing procedure
The Google Pixel 6 Pro's 50-megapixel main camera is joined by a 48MP telephoto camera with 4x optical zoom, plus a 12MP wide-angle camera. How about the Galaxy S21 Ultra? It was released in January 2021, but still has an enviable specification with a 108MP main camera, a 12MP wide-angle, a 10MP telephoto camera for 3x optical zoom photos, and a 10MP periscope camera for 10x optical zoom photos.

For this comparison, photos were taken back-to-back using the automatic mode, meaning all I did in most cases was press the shutter button. I tested the main camera, wide-angle, optical zoom, portrait, and night modes, but not video. All photos have been compared on a color-calibrated monitor, then resized for a friendlier online viewing experience. Do remember this when looking at the examples below.

Read more