Network Capacity Bound for Personalized PageRank in Multimodal Networks

In this paper, we introduce the Attention-Enhanced Multimodal Temporal Network (AMTN) to address the MuSe 2024 Cross-Cultural Humor Detection Sub-Challenge (MuSe-Humor), which highlights the task of humor detection within a cross-cultural context, leveraging multimodal information for its accomplishment. Specifically, we employ the Temporal Convolutional Network (TCN) to capture the temporal dynamics within individual modalities' features. Following this, we apply attention mechanism to refine the integration of information across different modalities and temporal sequences. The integrated features are then used for humor detection. Furthermore, we investigate the effectiveness of an end-to-end approach for this challenge. Finally, a more robust outcome is achieved by aggregating multiple experimental results, which constitutes our final submission for the challenge. As a result, our solution achieves a remarkable AUC score of 0.8833 on the test dataset, outperforming the baseline by 0.0151 and securing 2nd place in the MuSe-Humor sub-challenge.

Large-scale multimodal transportation network models and algorithms-Part II: Network capacity and network design problem

In the pursuit of capacity-oriented urban sustainable development (SD), this study aims to enhance the multimodal transport network by making it more accessible, efficient, and environmentally friendly. A sustainable framework combining policy mixes and second-best constraints is developed as a multimodal transportation network capacity (TNC) model with the bi-level programming (BLP) formulation. The upper-level model is to maximise the total origin-destination (OD) demand with considering second-best constraints. The lower-level model is a combined mode split and traffic assignment with elastic demand (CMSTA-ED) model which considers policy mixes. Numerical experiments are conducted to demonstrate the effectiveness of the hybrid PSO–GWO algorithm, the synergy of the method for selecting and mixing policy instruments based on the proposed assessment system, and a methodology to establish a policy mix. The findings of this research offer practical guidance for formulating well-informed and no-redundant policy mixes, contributing to the improvement of sustainable multimodal transportation network performance.

Catheter-based cardiac ablation is a minimally invasive procedure for treating atrial fibrillation (AF). Electrophysiologists perform the procedure under image guidance during which the contact force between the heart tissue and the catheter tip determines the quality of lesions created. This paper describes a novel multi-modal contact force estimator based on Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs). The estimator takes the shape and optical flow of the deflectable distal section as two modalities since frames and motion between frames complement each other to capture the long context in the video frames of the catheter. The angle between the tissue and the catheter tip is considered a complement of the extracted shape. The data acquisition platform measures the two-degrees-of-freedom contact force and video data as the catheter motion is constrained in the imaging plane. The images are captured via a camera that simulates single-view fluoroscopy for experimental purposes. In this sensor-free procedure, the features of the images and optical flow modalities are extracted through transfer learning. Long Short-Term Memory Networks (LSTMs) with a memory fusion network (MFN) are implemented to consider time dependency and hysteresis due to friction. The architecture integrates spatial and temporal networks. Late fusion with the concatenation of LSTMs, transformer decoders, and Gated Recurrent Units (GRUs) are implemented to verify the feasibility of the proposed network-based approach and its superiority over single-modality networks. The resulting mean absolute error, which accounted for only 2.84% of the total magnitude, was obtained by collecting data under more realistic circumstances in contrast to previous research studies. The decrease in error is considerably better than that achieved by individual modalities and late fusion with concatenation. These results emphasize the practicality and relevance of utilizing a multimodal network in real-world scenarios.

Estimation of urban network capacity with second-best constraints for multimodal transport systems

Emergency communication systems play a major role in disaster‐relief environments. In terms of the public safety research, the emergency relief communication system can provide a high system capacity for networks based on the development of Long‐Term Evolution. However, in the event of a disaster, mass traffic information can cause congestion in the core network, and communications between relief workers may be limited. Consequently, spectrum efficiency can be very weak. This paper provides a hybrid traffic offloading mechanism combining Device‐to‐Device (D2D) and Local IP Access (LIPA) modes for the disaster‐resilient network. With receiving power, the distance between relief workers and the distance between relief workers and the vehicular eNodeB (VeNB) as the LIPA/D2D switching criteria, the network can select an appropriate mode to prevent core network congestion. This paper also considers the effects of the mobility models (i.e., random walk and random direction) on the spectrum efficiency of the disaster‐resilient communication system. The proposed hybrid LIPA/D2D traffic offloading mechanism can prevent the local communication traffic from flowing into the core network and significantly improve the system spectrum efficiency when the core network is under congestion. Therefore, the proposed mechanism can effectively improve the quality of the communication between relief workers served by the same VeNB for performing rescue operations. Moreover, the hybrid LIPA/D2D traffic offloading mechanism can be applied to the smart city and smart home in the future.

Optical transport network (OTN) and software-defined networking (SDN) are widely used in backbone and metropolitan transmission networks, to improve network transmission capacity. In OTN, it is particularly important to correctly allocate routes and maximize network capacity. In order to solve this problem, this paper designs a random walk algorithm which is offline routing strategy based on Q-learning to optimize the network routing of OTN. The routing strategy uses traffic demand as the reward to drive the decision body, so that the decision body randomly explores the network topology, learns the current network state information, and makes a better routing decision. We model and experiment with the SD-OTN scenarios of this paper. The experimental results show that the routing strategy of the network architecture of this paper can achieve superior performance.