Abstract: An electrocardiogram detection device whose housing may be made of a conductive material, and the electrocardiogram detection device may include a voltage holder circuit configured to provide a target potential for the housing. A potential difference between the target potential provided by the voltage holder circuit and a reference potential provided by an electrocardiogram detection circuit for a third electrode is small. Therefore, in an ECG detection process, even if a user accidentally touches the housing and causes the housing to be conducted to the third electrode, no leakage current is generated between the housing and the third electrode, or a small leakage current is generated between the housing and the third electrode. This can effectively reduce interference to an ECG signal and ensure accuracy of ECG detection.

Abstract: An image display method and a related apparatus provided in embodiments of this application are applied to the field of terminal technologies. A first application and a second application are applications running in a foreground of an electronic device, and a first rendering priority of the first application is determined based on a first initial rendering priority of the first application, whether the first application corresponds to a focus window, and/or an application switching type of the first application. A second rendering priority of the second application is determined based on a second initial rendering priority of the second application, whether the second application corresponds to a focus window, and/or an application switching type of the second application.

Abstract: An electrocardiogram detection device (000) whose housing (10) may be made of a conductive material, and the electrocardiogram detection device (000) may include a voltage holder circuit (30) configured to provide a target potential for the housing (10). A potential difference between the target potential provided by the voltage holder circuit (30) and a reference potential provided by an electrocardiogram detection circuit (20) for a third electrode (P3) is small. Therefore, in an ECG detection process, even if a user accidentally touches the housing (10) and causes the housing (10) to be conducted to the third electrode (P3), no leakage current is generated between the housing (10) and the third electrode (P3), or a small leakage current is generated between the housing (10) and the third electrode (P3). This can effectively reduce interference to an ECG signal and ensure accuracy of ECG detection.

Abstract: Noise and channel distortion parameters in the vectorized logarithmic or the cepstral domain for an utterance may be estimated, and subsequently the distorted speech parameters in the same domain may be updated using an unscented transformation framework during online automatic speech recognition. An utterance, including speech generated from a transmission source for delivery to a receiver, may be received by a computing device. The computing device may execute instructions for applying the unscented transformation framework to speech feature vectors, representative of the speech, in order to estimate, in a sequential or online manner, static noise and channel distortion parameters and dynamic noise distortion parameters in the unscented transformation framework. The static and dynamic parameters for the distorted speech in the utterance may then be updated from clean speech parameters and the noise and channel distortion parameters using non-linear mapping.

Abstract: Noise and channel distortion parameters in the vectorized logarithmic or the cepstral domain for an utterance may be estimated, and subsequently the distorted speech parameters in the same domain may be updated using an unscented transformation framework during online automatic speech recognition. An utterance, including speech generated from a transmission source for delivery to a receiver, may be received by a computing device. The computing device may execute instructions for applying the unscented transformation framework to speech feature vectors, representative of the speech, in order to estimate, in a sequential or online manner, static noise and channel distortion parameters and dynamic noise distortion parameters in the unscented transformation framework. The static and dynamic parameters for the distorted speech in the utterance may then be updated from clean speech parameters and the noise and channel distortion parameters using non-linear mapping.