Abstract: A device includes a first stage keyword detector and a processor. The processor is configured to, based on a signal indicating a keyword detection by the first stage keyword detector and prior to confirmation of the keyword detection, initiate a change of state to enable a voice service that is in a low-power mode to be scheduled for execution on one or more reserved processor cores.

Abstract: A method for correcting PET data includes acquiring first PET data at a time interval. The method also includes acquiring a first normalization coefficient corresponding to the first PET data. The method also includes determining a scale factor based at least partially on the first normalization coefficient. The method also includes determining second PET data based on the first PET data, the scale factor, and the second normalization coefficient. The method also includes determining a first dead time correction coefficient corresponding to the second PET data. The method also includes determining third PET data based on the second PET data and the first dead time correction coefficient. The method further includes reconstructing a first image based on the third PET data.

Abstract: The present disclosure relates to a method and apparatus for processing image. The method includes: determining a first translation amount based on a first voice coil motor (VCM) value of a first image photographed by a wide-angle camera, the VCM value is a normalized value corresponding to a magnitude of a current acting on a coil of a VCM in response to successful focusing of the camera; determining a second translation amount based on a zoom multiple of the first image and the first translation amount; and obtaining a target image by translating the first image by the second translation amount and amplifying the first image. According to the present disclosure, smooth switching of zooming can be achieved, and obvious geometric jump of images in continuous focal lengths and between focal lengths is avoided, thereby providing a good zooming experience to the user.

Abstract: Embodiments include methods for voice/speech recognition in noisy environments executed by a processor of a computing device. In various embodiments, voice or speech recognition may be executed by a processor of a computing device, which may include determining a voice recognition model to use for voice and/or speech recognition based on a location where an audio input is received and performing voice and/or speech recognition on the audio input using the determined voice recognition model. Some embodiments my receive from a computing device, an audio input and location information associated with a location where the audio input was recorded. The received audio input may be used to generate a voice recognition model associated with the location where the audio input was recorded for use in voice and/or speech recognition. The generated voice recognition model associated with the location may be provided to the computing device.

Abstract: Various embodiments include methods for preloading data onto a wireless device based on predicted network status coverage and predicted user data access behaviors. In an aspect, a server may receive trip information a wireless device, receive network coverage map information representing a different network coverage areas, predict a travel time for the wireless device to reach the network coverage area with reduced signal quality, predict a travel duration of the wireless device within that network coverage area, and transmit a notification to the wireless device to enable preloading appropriate data before entering the network coverage area.

Abstract: The present disclosure relates to a method and apparatus for processing image. The method includes: determining a first translation amount based on a first voice coil motor (VCM) value of a first image photographed by a wide-angle camera, the VCM value is a normalized value corresponding to a magnitude of a current acting on a coil of a VCM in response to successful focusing of the camera; determining a second translation amount based on a zoom multiple of the first image and the first translation amount; and obtaining a target image by translating the first image by the second translation amount and amplifying the first image. According to the present disclosure, smooth switching of zooming can be achieved, and obvious geometric jump of images in continuous focal lengths and between focal lengths is avoided, thereby providing a good zooming experience to the user.

Abstract: A method for correcting PET data includes acquiring first PET data at a time interval. The method also includes acquiring a first normalization coefficient corresponding to the first PET data. The method also includes determining a scale factor based at least partially on the first normalization coefficient. The method also includes determining second PET data based on the first PET data, the scale factor, and the second normalization coefficient. The method also includes determining a first dead time correction coefficient corresponding to the second PET data. The method also includes determining third PET data based on the second PET data and the first dead time correction coefficient. The method further includes reconstructing a first image based on the third PET data.

Abstract: A method for correcting PET data includes acquiring first PET data at a time interval. The method also includes acquiring a first normalization coefficient corresponding to the first PET data. The method also includes determining a scale factor based at least partially on the first normalization coefficient. The method also includes determining second PET data based on the first PET data, the scale factor, and the second normalization coefficient. The method also includes determining a first dead time correction coefficient corresponding to the second PET data. The method also includes determining third PET data based on the second PET data and the first dead time correction coefficient. The method further includes reconstructing a first image based on the third PET data.

Abstract: A method for correcting PET data includes acquiring first PET data at a time interval. The method also includes acquiring a first normalization coefficient corresponding to the first PET data. The method also includes determining a scale factor based at least partially on the first normalization coefficient. The method also includes determining second PET data based on the first PET data, the scale factor, and the second normalization coefficient. The method also includes determining a first dead time correction coefficient corresponding to the second PET data. The method also includes determining third PET data based on the second PET data and the first dead time correction coefficient. The method further includes reconstructing a first image based on the third PET data.

Abstract: A method for correcting PET data includes acquiring first PET data at a time interval. The method also includes acquiring a first normalization coefficient corresponding to the first PET data. The method also includes determining a scale factor based at least partially on the first normalization coefficient. The method also includes determining second PET data based on the first PET data, the scale factor, and the second normalization coefficient. The method also includes determining a first dead time correction coefficient corresponding to the second PET data. The method also includes determining third PET data based on the second PET data and the first dead time correction coefficient. The method further includes reconstructing a first image based on the third PET data.

Abstract: A method for correcting PET data includes acquiring first PET data at a time interval. The method also includes acquiring a first normalization coefficient corresponding to the first PET data. The method also includes determining a scale factor based at least partially on the first normalization coefficient. The method also includes determining second PET data based on the first PET data, the scale factor, and the second normalization coefficient. The method also includes determining a first dead time correction coefficient corresponding to the second PET data. The method also includes determining third PET data based on the second PET data and the first dead time correction coefficient. The method further includes reconstructing a first image based on the third PET data.