Abstract: The present disclosure provides a method for multimodal emotion classification based on modal space assimilation and contrastive learning. The present disclosure introduces the concept of assimilation. A guidance vector composed of complementary information between modalities is utilized to guide each modality to simultaneously approach a solution space. This operation not only further improves the efficiency of searching for the solution space but also renders heterogeneous spaces of three modalities isomorphic. In a process of making spaces isomorphic, contributions of a plurality of modalities to a final solution space can be effectively balanced to a certain extent. When guiding each modality, this strategy enables a model to be more concerned about emotion features, thereby reducing intra-modal redundancy. Thus, the difficulty of establishing a multimodal representation is reduced.

Abstract: A neural network optimization method includes receiving a model file of a to-be-optimized neural network; obtaining a search space of a target neural network architecture based on the model file of the to-be-optimized neural network, where the search space includes a value range of each attribute of each neuron in the target neural network architecture; obtaining the target neural network architecture based on the search space; training the target neural network architecture based on the model file of the to-be-optimized neural network, to obtain a model file of a target neural network; and providing the model file of the target neural network to a user.

Abstract: System and method embodiments are provided for sparse beamforming design. In an embodiment, a method of designing sparse transmit beamforming for a network multiple-input multiple output (MIMO) system includes dynamically forming, by a cloud central processor, a cluster of transmission points (TPs) for use in transmit beamforming for each of a plurality of user equipment (UEs) in the system by optimizing a network utility function and system resources; determining, by the cloud central processor, a sparse beamforming vector for each UE according to the optimizing; and transmitting, by the cloud central processor, a message and first beamforming coefficients to each TP in the formed cluster associated with a first UE in the plurality of UEs, wherein each TP in the formed cluster associated with the first UE correspond to nonzero entries in a first beamforming vector corresponding to the first UE.