OnePlus might introduce AI enhancements in Camera app with Android Q
When it comes to the camera department, OnePlus is overshadowed by OEMs like Google and Huawei. While these companies extensively rely on hardware based computational photography to enhance the final output, OnePlus is having their faith on traditional software oriented post processing methods.
While OnePlus 7 Pro, the latest flagship from the Chinese OEM, does hold a pretty high position in the DxOMark ranking, regular users failed to replicate the said level of performance on their units.
OnePlus did release an array of software updates – a couple of them were specifically intended for camera optimizations – to make it right. Problem is, they still don’t want to infuse the power of so-called ‘artificial intelligence’ like other companies.
Our readers should remember the incident, when OnePlus imaging director Simon Liu was interviewed by Wired weeks before the official release of OnePlus 7 Pro. He emphasized on the pixel-binning capabilities of the Sony IMX586 sensor and the wide angle lens, but was reluctant when asked about introducing AI.
The OnePlus 7’s main camera, in contrast, uses downsampling or pixel binning to create 12-megapixel images and in the software version we used, there was no option to create photos at “full” resolution.
We slightly change the tuning but OnePlus users shouldn’t expect too much difference on the surface. If you’re trying to shoot a plant and the camera tells you it’s a dog, that’s not useful.
Although we don’t want to dive into the controversy about “AI versus a bunch of if-else clauses”, a recent interview of Zack Zhang, OnePlus Product manager, explores the shortcomings of camera processing by the company.
Conducted by The Next Web (TNW), Zhang did mention their plan to give more focus on computational photography as one of the key measures to stay strong in the current competition.
Even by our own standards, the camera is not where we want it to be. We’re trying to improve our image quality to match the best in the market.
Call it a coincidence, but we are able to spot an interesting development inside the stock OnePlus Camera app. Thanks to the findings of XDA forum member Bartixxx, we are now pretty much confident that the Shenzhen based OEM is actively working on integrating AI related features in their camera.
Noticed anything out of the place? The snippet is part of a difference between the strings found after decompiling different version of OnePlus Camera apps. To be specific, the 3.8.x branch comes with a reference of ‘Snapdragon Neural Processing Engine’ (look at line 439)!
FYI, Qualcomm publicly provides documentations for devs to unleash the true potential of their custom Kryo CPU cores, Adreno GPU and Hexagon DSPs. Using the aforementioned API, third party applications can use the AI functionalities of these components in a streamlined manner.
Premium tier Qualcomm® Snapdragon™ mobile platforms have extensive heterogeneous computing capabilities that are engineered to allow the running of trained neural networks on device without a need for connection to the cloud. The Qualcomm® Neural Processing SDK for artificial intelligence (AI) is designed to help developers run one or more neural network models trained in Caffe/Caffe2, ONNX, or TensorFlow on Snapdragon mobile platforms, whether that is the CPU, GPU or DSP.
The 3.8.x branch of OnePlus Camera is only available in the recently published Android Q Developer Preview 4, although it is possible to sideload the APK on older models.
We still don’t know how the company is planning to utilize the AI modules to strengthen their camera processing, but guess it should be publicly available with the stable release of Android Q.
PiunikaWeb is a unique initiative that mainly focuses on investigative journalism. This means we do a lot of hard work to come up with news stories that are either ‘exclusive,’ ‘breaking,’ or ‘curated’ in nature. Perhaps that’s the reason our work has been picked by the likes of Forbes, Foxnews, Gizmodo, TechCrunch, Engadget, The Verge, Macrumors, and more. Do take a tour of our website to get a feel of our work. And if you like what we do, stay connected with us on Twitter (@PiunikaWeb) and other social media channels to receive timely updates on stories we publish.
Talk to a JustAnswer expert if you have any related query:
DISCLAIMER: PiunikaWeb & JustAnswer are in a paid collaboration, which means a reader's successful engagement with the JustAnswer service will result in some commission being passed on to PiunikaWeb.