Xiaomi Imaging Brain On Xiaomi 12 Series
Today Xiaomi published an official article detailing the imaging system of the Xiaomi 12 Series. The Xiaomi 12 Series focuses on solving the problem of not taking pictures fast enough in the past, while the Xiaomi 12 Pro debuts the Sony IMX707 sensor, which has better night scene shooting capabilities.
Xiaomi said that it took two years to rewrite the entire camera architecture, and rebuild the integration of hardware and software capabilities to create a new Xiaomi imaging brain, and Xiaomi 12 Series equipped with Xiaomi Imaging Brain, containing a total of four parts: intention recognition unit, computing unit, acceleration engine, ecological engine.
- The role of the intent recognition unit is to determine the user’s intent and then dynamically adjust the intensity of the use of algorithms, including intelligent dynamic allocation of memory and so on. When the user keeps pressing the shutter, at this time Xiaomi Image Brain will think that the user intends to record more images, and then it will give priority to ensure the speed of taking pictures.
- The acceleration engine is empowered by the Snapdragon 8 Gen1 platform, which integrates and fully schedules the computing power of CPU, GPU, DSP, ISP, and other modules, allowing the whole process to work in parallel, thus achieving a significant speed optimization effect.
- The computation unit is a powerful computational photography capability that invokes advanced algorithms for image quality tuning, such as multi-frame compositing, noise reduction algorithms, artifact removal algorithms, white balance calibration, etc., to make the images clearer and more natural, thus achieving effects close to or even beyond traditional optical photography.
- The ecological engine opens up the pipeline from hardware to third-party applications, opening up the capabilities of Xiaomi Image Brain to third parties through Xiaomi’s SDK, so that third-party applications can also call on Xiaomi Image Brain’s capabilities in the future, thus achieving the effect of image quality tuning and speed improvement.
Officially, the Xiaomi 12 series camera starts 21.2% faster, takes photos 53.7% faster, takes continuous photos 209% faster, and takes night shots 39.9% faster. The Xiaomi Imagine Brain’s heterogeneous computing and algorithmic parallel processing play a key role in this.
In the past photo process, each photo from pressing the shutter, to the algorithm processing, and finally in the album to generate a preview, in the middle of every place is serial computing, which means that the current photo is completely processed before the completion of the calculation, is unable to respond to the next photo, reflected in the photo experience is a continuous shooting shutter often can not press the situation, often want to capture a few more, but repeatedly miss the best time.
Xiaomi Imaging Brain adopts a new parallel processing method, allowing the processing algorithm to move up and process multiple photos in parallel. Then the original hardware HAL layer is abstracted and managed as Xiaomi’s own HAL layer, allowing the hardware algorithm to be so combined and changed, and even plug-in management. Together with algorithm pre-loading optimization and algorithm concurrency optimization, the ability to process multiple consecutive photos is greatly improved. Heterogeneous computing can reduce unnecessary algorithm library loading, which results in lower memory and faster speed.
The first Sony ultra-sensitive outsole IMX707, 49% increase in light sensitivity to restore the pure night scene. With the development of cell phone cameras today, night scene shooting performance is still a proposition that has room for improvement. To bring better night scene performance, Xiaomi spared no cost to invest in the Xiaomi 12 Pro, the first to be equipped with the IMX707 Sony super bottom sensor. This CMOS sensor has a large sensing area of 1/1.28″, with a single-pixel size of 1.22 μm and 2.44 μm after the pixel quadruplex, providing stronger light sensitivity, and a 49% increase in light sensitivity compared to the IMX766, the “big cup” flagship favorite on the market.
The IMX707 not only makes the Xiaomi 12 Pro’s night scenes purer, but the larger sensing area also brings increased photo speed and enhanced capture ability. When the light intake is increased, the Xiaomi 12 Pro can get better brightness performance with a shorter exposure time, thus increasing the shutter speed and allowing for clearer captures to freeze beautiful moments.
Whether it is hardware or software, the ultimate return is the consumer experience, from rewriting the camera architecture to create Xiaomi imaging brain to the first equipped with IMX707, cell phone imaging capabilities to improve is not just a hardware race, but a combination of hard and soft multi-dimensional ability to compete, both hands should be grasped, both hands should also be hard. Xiaomi 12 series choose a new generation of sensors is also not because only of the new hardware to take better photos, but the new hardware and Xiaomi imaging brain capabilities of a better combination, to allow Xiaomi imaging brain to play the maximum effect, to bring the unprecedented image experience.
