Abstract

The content of notifications from the device operating system and applications is displayed on the screen of a device even when the device is locked. Such notifications make private or sensitive information accessible by other parties without unlocking the device. To counter this risk, device operating systems allow users to hide notification details when the device is locked. However, such hiding results in significant loss of understandability and utility of notifications. Per techniques of this disclosure, a trained machine learning based model is applied with specific user permission to identify sensitive content by analyzing pixels of the device lock screen image that includes notifications. The techniques may be incorporated within the device operating system or web application frameworks. If the inferred level of sensitivity is higher than a threshold value, the corresponding content is obfuscated when displaying the notification on a locked device.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.

Share

COinS