Textbook question answering is challenging as it aims to automatically answer various questions on textbook lessons with long text and complex diagrams, requiring reasoning across modalities. In this work, we propose MRHF, a novel framework that incorporates dense passage re-ranking and the mixture-of-experts architecture for TQA. MRHF proposes a novel query augmentation method for diagram questions and then adopts multi-stage dense passage re-ranking with large pretrained retrievers for retrieving paragraph-level contexts. Then it employs a unified question solver to process different types of text questions. Considering the rich blobs and relation knowledge contained in diagrams, we propose to perform multimodal feature fusion over the retrieved context and the heterogeneous diagram features. Furthermore, we introduce the mixture-of-experts architecture to solve the diagram questions to learn from both the rich text context and the complex diagrams and mitigate the possible negative effects between features of the two modalities. We test the framework on the CK12-TQA benchmark dataset, and the results show that MRHF outperforms the state-of-the-art results in all types of questions. The ablation and case study also demonstrates the effectiveness of each component of the framework.

@en

China will launch the “Tianwen-IV” mission around 2030, focusing on the orbiting exploration of Jupiter and Callisto, a moon of Jupiter. As part of this ambitious mission, a main satellite will carry another satellite that will be released in the Jupiter system to continue its journey toward Uranus. Considering the current mission planning, we propose an inter-satellite radio-observation mode that differs from the conventional observation mode of tracking from Earth to precisely determine the orbit of the satellites. Given the significance of the Callisto gravity field model in both science objectives and satellite navigation, we have conducted a series of simulation experiments to evaluate the potential of this inter-satellite range-rate data for accurately estimating the Callisto gravity field. The results obtained from the analysis demonstrate that by utilizing 40 days of ground station observations, it is possible to estimate the gravity field model of Callisto up to a degree of 70. Remarkably, when combining these ground station observations with inter-satellite observations, a comparable level of accuracy can be achieved with just 10 days of observations. Furthermore, with reduced inter-satellite observation noise, accuracy improves, enabling estimation up to 80 degrees or higher. Initial inter-satellite distance selection impacts estimation accuracy. These findings serve as a valuable test bed for the future “Tianwen-IV” mission to perform precise orbit determination and gravity field model estimation to reduce reliance on deep space stations.

@en

Cross-modal retrieval, as an important emerging foundational information retrieval task, benefits from recent advances in multimodal technologies. However, current cross-modal retrieval methods mainly focus on the interaction between textual information and 2D images, lacking research on 3D data, especially point clouds at scene level, despite the increasing role point clouds play in daily life. Therefore, in this paper, we proposed a cross-modal point cloud retrieval benchmark that focuses on using text or images to retrieve point clouds of indoor scenes. Given the high cost of obtaining point cloud compared to text and images, we first designed a pipeline to automatically generate a large number of indoor scenes and their corresponding scene graphs. Based on this pipeline, we collected a balanced dataset called CRISP, which contains 10K point cloud scenes along with their corresponding scene images and descriptions. We then used state-of-the-art models to design baseline methods on CRISP. Our experiments demonstrated that point cloud retrieval accuracy is much lower than cross-modal retrieval of 2D images, especially for textual queries. Furthermore, we proposed ModalBlender, a tri-modal framework which can greatly improve the Text-PointCloud retrieval performance. Through extensive experiments, CRISP proved to be a valuable dataset and worth researching. (Dataset can be downloaded at https://github.com/CRISPdataset/CRISP.)

@en

The synchronization of power generators is an important condition for the proper functioning of a power system, in which the fluctuations in frequency and the phase angle differences between the generators are sufficiently small when subjected to stochastic disturbances. Serious fluctuations can prompt desynchronization, which may lead to widespread power outages. Here, we model the stochastic disturbance by a Brownian motion process in the linearized system of the non-linear power systems and characterize the fluctuations by the variances of the frequency and the phase angle differences in the invariant probability distribution. We propose a method to calculate the variances of the frequency and the phase angle differences. For the system with uniform disturbance-damping ratio, we derive explicit formulas for the variance matrices of the frequency and the phase angle differences. It is shown that the fluctuation of the frequency at a node depends on the disturbance-damping ratio and the inertia at this node only, and the fluctuations of the phase angle differences in the lines are independent of the inertia. In particular, the synchronization stability is related to the cycle space of the network. We reveal the influences of constructing new lines and increasing capacities of lines on the fluctuations in the phase angle differences in the existing lines. The results are illustrated for the transmission system of Shandong Province of China. For the system with non-uniform disturbance-damping ratio, we further obtain bounds of the variance matrices.@en

The synchronization stability of a complex network system of coupled phase oscillators is discussed. In case the network is affected by disturbances, a stochastic linearized system of the coupled phase oscillators may be used to determine the fluctuations of phase differences in the lines between the nodes and to identify the vulnerable lines that may lead to desynchronization. The main result is the derivation of the asymptotic variance matrices of the phase differences which characterizes the severity of the fluctuations. It is found that the cycle space of the graph of the system plays a role in this characterization. With theory of the cycle space, the effect of forming small cycles on the fluctuations is evaluated. It is proven that adding a new line or increasing the coupling strength of a line affects the fluctuations in the lines in any cycle including this line, while it does not affect the fluctuations in the other lines. In particular, if the phase differences at the synchronous state are not changed by these actions, then the affected fluctuations reduce.

@en

Current news datasets merely focus on text features on the news and rarely leverage the feature of images, excluding numerous essential features for news classification. In this paper, we propose a new dataset, N24News, which is generated from New York Times with 24 categories and contains both text and image information in each news. We use a multitask multimodal method and the experimental results show multimodal news classification performs better than text-only news classification. Depending on the length of the text, the classification accuracy can be increased by up to 8.11%. Our research reveals the relationship between the performance of a multimodal classifier and its sub-classifiers, and also the possible improvements when applying multimodal in news classification. N24News is shown to have great potential to prompt the multimodal news studies.

@en

Due to low power consumption, long lifetime and many other advantages, Light-emitting Diode (LED) has increased dramatically all over the world. Chip Scale Package (CSP) LED is a new LED package with small size, high current, and high reliability. For the CSP LED with smaller size and lower thermal resistance, phosphors and silicone are usually combined as the phosphors/silicone composite and prepared by using a high temperature cure process. However, for those CSPs which are not sufficiently cured, their reliability under a harsh environment (e.g. high temperatures and high humidity) will obviously decrease. In this study, the influence of temperature and ultraviolet light on the cure process of phosphor/silicone composites is studied and an optimal cure process is extracted accordingly. According to the cure experiment under different conditions, the results show that ultraviolet light and phosphor can promote the cure reaction of silicone. With the increase of ultraviolet light intensity and phosphor mass fraction, the cure rate of the silicone and phosphor/silicone composite is greatly increased as well.

@en