Sony Ericsson Applies for Object Recognition Patent
Sony Ericsson has applied for a patent that would allow camera-equipped cell phones to take pictures or videos of objects and then use on-board pattern recognition software to determine what the object is. It would display text on the screen with the name of the object. The patent also says that if he software can't determine the exact name of the object, that it at least be able to classify the object as part of a group. The patent includes the capability to use multiple languages, store object names in the phone's memory, and geotag objects captured and identified. Sony Ericsson first applied for the patent in June 2007. Sony Ericsson has not said if/when it would include such a feature in its cell phones.
Google Lens Uses Machine Learning to Recognize Objects
Google today announced Google Lens, an image-recognition tool that relies on mobile cameras to perform searches. The tool is a significant advancement to the old Google Image Search app.
Google Says ARCore Will Push Augmented Reality to the Android Masses
Google today introduced ARCore, a software developer kit that will bring augmented reality to existing and future Android smartphones without the special hardware required by Project Tango. Google says it built ARCore on the foundation set by Project Tango, but has tweaked it in a way the will allow most devices to run AR apps and content.
Google Partners Create Add-Ons for Docs and Sheets
Google today said it has worked with several partners to develop Android add-ons for its Docs and Sheets productivity apps. The add-ons extend the functionality of Docs and Sheets, allowing people to accomplish more from their mobile phones.
Google to Bring Object Recognition to Phones
Google today said it has partnered with semiconductor startup Movidius to bring deep learning to smartphones. Movidius specializes in computer vision and helped with Google's Project Tango 3D mapping effort.