Abstract: Systems, devices, and techniques are provided for reducing graphical bandwidth in a projection display system by modifying a modulation frequency used to display one or more portions of content based on an identified content type of such portions. Content is received for display, and types of content associated with one or more portions of the received content are identified. For each respective portion of content, an effective resolution is selected for displaying the respective portion based on the identified content type for the respective portion, and one or more light beams directed to a respective display area of a projection surface are modulated at a respective emitter modulation frequency that corresponds to the selected effective resolution for the respective portion.

Abstract: A method for rejecting biased data using a machine learning model includes receiving a cluster training data set including a known unbiased population of data and training a clustering model to segment the received cluster training data set into clusters based on data characteristics of the known unbiased population of data. Each cluster of the cluster training data set includes a cluster weight. The method also includes receiving a training data set for a machine learning model and generating training data set weights corresponding to the training data set for the machine learning model based on the clustering model. The method also includes adjusting each training data set weight of the training data set weights to match a respective cluster weight and providing the adjusted training data set to the machine learning model as an unbiased training data set.

Abstract: In one aspect, this application describes a method for determining a version of a software application targeted for a computing device. The method includes receiving, at an application marketplace system and from a user associated with a computing device that operates remotely from the application marketplace system, a request that corresponds to a software application distributed by the application marketplace system, the software application having multiple versions on the application marketplace system. The method also includes determining one or more device attributes that are associated with the computing device, and identifying a particular version of the software application, from among the multiple versions on the application marketplace system, that is targeted for the computing device based on the device attributes. The method also includes providing, for display to the user and in response to the request, information related to the particular version of the software application.

Abstract: Systems, methods, and computer storage mediums are provided for collaboratively editing geographic features in a shared geographic database. An example method includes receiving a request from a first user to edit data associated with a plurality of geographic features geocoded within a first geographic area. The first geographic area is selected by the first user and defined by a first set of coordinates. The method then determines whether the first geographic area intersects a second geographic area. The second geographic area is selected by a second user and defined by a second set of coordinates. The intersection between the first and second sets of coordinates is determined by comparing the coordinates of the first and second geographic areas. When the first and second geographic areas do not intersect, the method locks the first geographic area to prevent other users from editing the geographic features geolocated within the first geographic area.

Abstract: The subject matter of this specification can be embodied in, among other things, a computer-implemented method for removing noise from audio that includes building a sound model that represents noises which result from activations of input controls of a computer device. The method further includes receiving an audio signal produced from a microphone substantially near the computer device. The method further includes identifying, without using the microphone, an activation of at least one input control from among the input controls. The method further includes associating a portion of the audio signal as corresponding to the identified activation. The method further includes applying, from the audio model, a representation of a noise for the identified activation to the associated portion of the audio signal so as to cancel at least part of the noise from the audio signal.