Abstract: The present disclosure is directed to methods and apparatus for tuning implementations of machine learning models for resource-constrained devices. In various embodiments, computer-readable instructions that utilize a trained machine learning model during execution of the computer-readable instructions on a resource-constrained device may be evaluated. Based on the evaluating, it may be determined that an amount of a first computing resource required by the resource-constrained device to execute the computer-readable instructions fails to satisfy a constraint related to the first computing resource. Based on the constraint, one or more candidate alterations to the computer-readable instructions may be identified.

Abstract: The present disclosure is directed to methods and apparatus for evaluating resources that would be used by machine learning model(s) for purposes of implementing the machine learning model(s) on resource-constrained devices. For example, in one aspect, a plurality of layers in a machine learning model may be identified. A plurality of respective output sizes corresponding to the plurality of layers may be calculated. Based on the plurality of output sizes, a maximum amount of volatile memory used for application of the machine learning model may be estimated and compared to a volatile memory constraint of a resource-constrained computing device. Output indicative of a result of the comparing may be provided at one or more output components.

Abstract: Disclosed herein is a medical instrument (100, 300, 400, 500) comprising: a memory (110) storing machine executable instructions (120) and a tomographic data assessment module (122) and a processor (106) configured for controlling the medical instrument. Execution of the machine executable instructions causes the processor to receive (200) measured tomographic data (124). The measured tomographic data is configured for being reconstructed into a tomographic image (308) of a subject (418). Execution of the machine executable instructions further causes the processor to receive (202) an image quality indicator (126, 126?, 126?) by inputting the measured tomographic data into the tomographic data assessment module. The tomographic data assessment module is configured for generating the image quality indicator in response to inputting the measured tomographic data.

Abstract: The present disclosure is directed to methods and apparatus for evaluating resources that would be used by machine learning model(s) for purposes of implementing the machine learning model(s) on resource-constrained devices. For example, in one aspect, a plurality of layers in a machine learning model may be identified. A plurality of respective output sizes corresponding to the plurality of layers may be calculated. Based on the plurality of output sizes, a maximum amount of volatile memory used for application of the machine learning model may be estimated and compared to a volatile memory constraint of a resource-constrained computing device. Output indicative of a result of the comparing may be provided at one or more output components.

Abstract: The present disclosure is directed to methods and apparatus for tuning implementations of machine learning models for resource-constrained devices. In various embodiments, computer-readable instructions that utilize a trained machine learning model during execution of the computer-readable instructions on a resource-constrained device may be evaluated. Based on the evaluating, it may be determined that an amount of a first computing resource required by the resource-constrained device to execute the computer-readable instructions fails to satisfy a constraint related to the first computing resource. Based on the constraint, one or more candidate alterations to the computer-readable instructions may be identified.

Abstract: A wearable sensor device which comprises a heart rate monitor, a motion sensor and an optical sensor. The heart rate monitor output, the motion sensor output and the optical sensor output are processed to determine if the sensor device is currently being worn. In this way, a robust mechanism is provided for detecting when the sensor is being worn. Monitoring for the next transition from a not worn state back to a worn state is achieved with lowest possible power consumption by using only the optical sensing.

Abstract: In an OFDM mobile communications system, an algorithm for forming a preliminary estimate of the channel on pilot subcarriers is carried out. Based on this preliminary estimate, a channel property is estimated. This channel property is then used to decide whether to enter a mobile receiver mode or a stationary receiver mode. For example, in the mobile mode channel estimation is performed using only pilot symbols from the current symbol period, while in the static mode channel estimation is performed using pilot symbols from the current symbol period and pilot symbols from other symbol periods.

Abstract: In an optical communication method, an accurate common clock signal (GPS) is provided to a sender (10) and a receiver (20). In the sender, the common clock signal is received, a first clock signal (CLK1) is generated on the basis of said common clock signal, a carrier wave with a predetermined carrier frequency (f) is generated using said clock signal as timing reference, the carrier wave is modulated with a data signal, and the modulated carrier wave is transmitted using a light beam (13). In the receiver, the common clock signal is received, a reference signal having the same frequency (f) as the carrier frequency is generated on the basis of said common clock signal, said optical beam (13) is received, a detection signal is derived from the optical beam, the receiver is tuned to the predetermined carrier frequency (f), and the data signal is derived from the detection signal.