Abstract: There is provided techniques for determining INACTIVE eDRX configurations for UEs. The method is performed by a core network node. The method comprises performing an iterative learning process with the access network nodes to determine the INACTIVE eDRX configurations. The iterative learning process comprises transmitting a global parameter vector of a machine learning model to the access network nodes for determining the INACTIVE eDRX configurations of the UEs. The global parameter vector defines coefficients for the machine learning model for the current iteration. The iterative learning process comprises receiving local model parameter vectors with locally updated coefficients for the machine learning model from the access network nodes. The iterative learning comprises updating the global parameter vector for the next iteration as a function of all received local model parameter vectors until a stopping criterion is fulfilled.

Abstract: Systems and methods for implementing Discontinuous Reception (DRX) are provided. In some embodiments, a method performed by a User Equipment (UE) for implementing DRX includes: configuring a DRX scheme which is activated in certain conditions; implementing DRX at the UE during one or more of the group consisting of: from the time the UE receives an RRCRelease message; after the UE receives DL data; and until terminating a (MT-SDT) session. This can result in significant energy benefits at the UE end. Some embodiments enable DRX during the MT-SDT procedure. The advantage of the solution is reduced energy consumption by the UE by saving the energy spent by enabling DRX in the MT-SDT procedure.

Abstract: A system and method for determining sub-activities in videos and segmenting the videos is disclosed. The method includes extracting one or more batches from one or more videos and extracting one or more features from set of frames associated with the one or more batches. The method further includes generating a set of predicted codes and determining a cross-entropy loss, temporal coherence loss and a final loss. Further, the method includes categorizing the set of frames into one or more predefined clusters and generating one or more segmented videos based on the categorized set of frames, the determined final loss, and the set of predicted codes by using s activity determination-based ML model. The method includes outputting the generated one or more segmented videos on user interface screen of one or more electronic devices associated with one or more users.

Abstract: An inspection robot for inspecting gastrointestinal tracts or other lumen. The inspection robot is capable of locomotion and comprises: a body, a vibration actuator, and a plurality of resilient legs. The plurality of resilient legs are arranged to protrude outwardly and rearwardly from the body, with respect to a direction of locomotion. Each leg is coupled to the vibration actuator at a proximal end of the leg and the vibration actuator is operable to induce vibrations in a distal end of the leg, which serve, in use, to propel the body in the direction of locomotion by the distal end generating a pushing force against an external surface.

Abstract: There is provided techniques for determining INACTIVE eDRX configurations for UEs. The method is performed by a core network node. The method comprises performing an iterative learning process with the access network nodes to determine the INACTIVE eDRX configurations. The iterative learning process comprises transmitting a global parameter vector of a machine learning model to the access network nodes for determining the INACTIVE eDRX configurations of the UEs. The global parameter vector defines coefficients for the machine learning model for the current iteration. The iterative learning process comprises receiving local model parameter vectors with locally updated coefficients for the machine learning model from the access network nodes. The iterative learning comprises updating the global parameter vector for the next iteration as a function of all received local model parameter vectors until a stopping criterion is fulfilled.

Abstract: A system and method for implementing variable-precision matrix multiplications using a low-precision digit matrix multiplier is disclosed. The system enables multiplication of matrices of different dimensions by splitting the large matrix into fixed-size matrix blocks. These block matrices are further decomposed into fixed-precision digit submatrices that are then individually multiplied, scaled, and accumulated to allow for variable-precision matrix multiplication. The system uses a systolic array of block matrix multipliers, which are each an array of dot product units, to efficiently implement larger matrix multiplications without substantially increasing either latency or wiring congestion. The system further uses only unsigned digit matrix multipliers but accounts for signed matrix multiplication by using row and column sums of the input matrices to adjust for the signed to unsigned conversion.

Abstract: A system and method for determining sub-activities in videos and segmenting the videos is disclosed. The method includes extracting one or more batches from one or more videos and extracting one or more features from set of frames associated with the one or more batches. The method further includes generating a set of predicted codes and determining a cross-entropy loss, temporal coherence loss and a final loss. Further, the method includes categorizing the set of frames into one or more predefined clusters and generating one or more segmented videos based on the categorized set of frames, the determined final loss, and the set of predicted codes by using s activity determination-based ML model. The method includes outputting the generated one or more segmented videos on user interface screen of one or more electronic devices associated with one or more users.