Nokia Technologies Oy

Joachim Wabnig

Antti Niskanen

David Bitauld

Hongwei Li

Miska Matias Hannuksela

Emre Baris Aksu

Francesco Cricri

Chris Bower

Iraj Saniee

Hamed Rezazadegan Tavakoli

Troels Roennow

Honglei Zhang

Andrew Peter Matthews

Enrique Martin Lopez

Troels F Roennow

An apparatus includes at least one processor; and at least one non-transitory memory including computer program code; wherein the at least one memory and the computer program code are configured to, with the at least one processor, cause the apparatus at least to: estimate an importance of parameters of a neural network based on a graph diffusion process over at least one layer of the neural network; determine the parameters of the neural network that are suitable for pruning or sparsification; remove neurons of the neural network to prune or sparsify the neural network; and provide at least one syntax element for signaling the pruned or sparsified neural network over a communication channel, wherein the at least one syntax element comprises at least one neural network representation syntax element.

Graph diffusion for structured pruning of neural networks

In accordance with an example embodiment a method, apparatus and computer program product are provided. The method comprises receiving, at an apparatus, a sampled data associated with an activity from one or more sensors wirelessly coupled to the apparatus. The sampled data is generated at the one or more sensors based on a compressive sampling of an activity data associated with the activity. The compressive sampling is performed based on a sampling information. The activity is classified based at least on the sampled data. An error associated with the classification of the activity is determined. The sampling information is updated or retained based on a comparison of the error with a threshold error. The updated sampling information is utilized for generating an updated sampled data. The updated sampled data facilitates in reclassification of the activity.

Method, apparatus and computer program product for activity recognition

A data processing system is disclosed for machine learning. The system comprises a sampling module () and a computational module () interconnected by a data communications link (). The computational module is configured to store a parameter vector representing an energy function of a network having a plurality of visible units connected using links to a plurality of hidden units, each link being a relationship between two units. The sampling module is configured to receive the parameter vector from the first processing module and to sample from the probability distribution defined by the parameter vector to produce state vectors for the network. The computational module is further configured to receive the state vectors from the second processing module and to apply an algorithm to produce new data. The sampling and computational modules are configured to operate independently from one another.

Data processing

Methods and apparatuses for communication between a mobile device and a target device are disclosed. Information of a target device is determined by means of at least one element of a mobile device for providing an optical link with the target device. An optical component of the mobile device is then aligned with an optical component of the target device based on said information determined by the mobile device. The target device can obtain information of relative positioning of the target device and the mobile device determined for the purposes of providing an optical link between the target device and the mobile device and the optical component thereof can be aligned with the optical component of the mobile device based on the information.

Optical link establishment

For secure wireless communications the sender device uses a rail encoder that outputs dual rail-encoded states of light in a time slot. The states of light dual rail-encode information according to a phase and/or intensity difference between the dual rails, and the rail-encoded states of light may further be converted to a polarization-encoded state. This may be implemented using at least two polarizing beam-splitters with at least one quarter-wave plate disposed therebetween; and/or with integrated waveguides that convert three optical inputs to two optical outputs that are input to a polarization rotator-combiner. The encoder may randomly define the polarization-encoded state such as by randomly selecting from a finite number of at least N=3 possible polarization rotations. The recipient device may use 2N parallel channels to decode the dual rail-encoded states of light, each channel comprising a detector configured to detect one of N possible polarization states.

Secured wireless communications

Distribution of a key in a wireless system comprising at least one mobile device is disclosed. Data regarding relative rotation between devices in optical data communications for key distribution is determined. The determined data is taken into account in determining how to use the optical data communications for establishing a key.

Key distribution in a wireless system

In accordance with an example embodiment of the present invention, there is provided an apparatus comprising a dual-rail encoder () configured to receive light from a light source and to output dual-rail encoded light, a combiner () configured to convert the dual-rail encoded light into polarization encoded light, and at least one processing core configured to obtain compensation adjustment information concerning a fiber () and to control the dual-rail encoder () based at least in part on the compensation adjustment information.

Fiber-based communication

Adjustment of an optical component of a device includes determining of at least one derivative of coupling efficiency of the optical component as a function of parameters used for control of a steering function of the optical component. At least one oscillating component is induced into the parameters for the determining. The adjustment of the optical component is based on the determined at least one derivative.

Alignment of optical devices

Methods and apparatus, including computer program products, are provided securing information. In some example embodiments, there is provided a method, which may include receiving, at a user equipment, biometric data from a user of the user equipment; receiving, at the user equipment, sensitive data to be encoded and/or decoded; encoding, based on at least one feature detected in the biometric data, the sensitive data, when the user equipment is in a storage mode; and decoding, based on at least the at least one feature detected in the biometric data, the sensitive data, when the user equipment is in a read mode. Related apparatus, systems, methods, and articles are also described.

Storing and retrieving cryptographic keys from biometric data

A system comprising: at least one sensor and at least one control apparatus wherein; the sensor comprises: at least one processor; and at least one memory including computer program code; the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus at least to perform; compressing a sensor data signal using a sampling basis to obtain a compressed data signal; and in response to a first feedback signal changing a sampling basis used to obtain the compressed data signal; and wherein the control apparatus comprises: at least one processor; and at least one memory including computer program code; the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus at least to perform; receiving the data signal from the at least one sensor; determining a quality of the received data signal; and if the quality of the received data signal is within a first threshold providing a feedback signal to control the sampling basis of the sensor.

Methods, apparatus and computer programs for obtaining data

Methods and apparatus for distribution of keys are disclosed. An optical signal for carrying encoded information in accordance with a quantum key distribution scheme is generated. The generated optical signal has a wavelength which is changed to another wavelength prior to transmission of the optical signal. The optical signal carrying the encoded information and having the changed wavelength is received, where after decoding of the information takes place by means of detector apparatus operating in the changed wavelength.

Quantum key distribution

The disclosure relates to secured wireless communications. A sender device emits randomly photons in a first polarization, a second polarization and a third polarization without aligning a polarization system in a plane perpendicular to the first and second polarizations with a recipient device. The recipient device is adapted for detection of events in association with six polarizations. The recipient device selects randomly polarization basis for measurement of the received photons. Information of detected events in association with three basis is communicated to the sender device. The information from the recipient device and information stored in the sender device is processed to determine events where same polarization basis was used by the sender device and the recipient device.

Growing community of inventors

Upper Cambourne, United Kingdom

Joachim Wabnig