While user preferences are important to cross-domain recommendation (CDR), existing methods primarily discover preferences under specific, yet possibly redundant, item features. To this end, we first propose a novel Preference Prototype-Aware (PPA) learning method to quantitatively learn user preferences while minimizing disturbances from the source domain. It introduces a mix-encoder and a proto-decoder. On the one hand, the mix-encoder learns better general representations of interacted items and captures the intrinsic relationships between items across different domains. On the other hand, the proto-decoder implements a learnable prototype matching mechanism to quantitatively perceive user preferences, avoiding disturbances caused by item features from the source domain. Moreover, through experiments on PPA, we observe another two issues that affect existing CDR methods’ performance, i.e., the semantic deficiency caused by sparse item categories and the imbalance weights caused by different user-item distributions. Thus, we further propose a LoRA-based extractor and a domain cross-attention module to alleviate the two issues, respectively. The PPA incorporating with new extractor and attention module is called PPA++. Extensive experiments show that PPA++ outperforms the other state-of-the-art counterparts in four different CDR scenarios.

Large language models (LLMs) have shown promise in multivariate time series classification (MTSC). To effectively adapt LLMs for MTSC, it is crucial to generate comprehensive and informative data representations. Most methods utilizing LLMs encode numerical time series into the model's latent space, aiming to align with the semantic space of LLMs for more effective learning. Despite effectiveness, we highlight three limitations that these methods overlook: (1) they struggle to incorporate temporal and channel-specific information, both of which are essential components of multivariate time series; (2) aligning the learned representation space with the semantic space of the LLMs proves to be a significant challenge; (3) they often require task-specific retraining, preventing training-free inference despite the generalization capabilities of LLMs. To bridge these gaps, we propose TableTime, which reformulates MTSC as a table understanding task. Specifically, TableTime introduces the following strategies: (1) utilizing tabular form to unify the format of time series, facilitating the transition from the model-centric approach to the data-centric approach; (2) representing time series in text format to facilitate seamless alignment with the semantic space of LLMs; (3) designing a knowledge-task dual-driven reasoning framework, TableTime, integrating contextual information and expert-level reasoning guidance to enhance LLMs' reasoning capabilities and enable training-free classification. Extensive experiments conducted on 10 publicly available benchmark datasets from the UEA archive validate the substantial potential of TableTime to be a new paradigm for MTSC. The code is publicly available. https://github.com/realwangjiahao/TableTime.

Large-scale Multi-Agent Reinforcement Learning (MARL) often suffers from the curse of dimensionality, as the exponential growth in agent interactions significantly increases computational complexity and impedes learning efficiency. To mitigate this, existing efforts that rely on Mean Field (MF) simplify the interaction landscape by approximating neighboring agents as a single mean agent, thus reducing overall complexity to pairwise interactions. However, these MF methods inevitably fail to account for individual differences, leading to aggregation noise caused by inaccurate iterative updates during MF learning. In this paper, we propose a Bi-level Mean Field (BMF) method to capture agent diversity with dynamic grouping in large-scale MARL, which can alleviate aggregation noise via bi-level interaction. Specifically, BMF introduces a dynamic group assignment module, which employs a Variational AutoEncoder (VAE) to learn the representations of agents, facilitating their dynamic grouping over time. Furthermore, we propose a bi-level interaction module to model both inter- and intra-group interactions for effective neighboring aggregation. Experiments across various tasks demonstrate that the proposed BMF yields results superior to the state-of-the-art methods. Our code is available at https://github.com/Chreer/BMF.

Pre-trained foundation models have recently made significant progress in table-related tasks such as table understanding and reasoning. However, recognizing the structure and content of unstructured tables using Vision Large Language Models (VLLMs) remains under-explored. To bridge this gap, we propose a benchmark based on a hierarchical design philosophy to evaluate the recognition capabilities of VLLMs in training-free scenarios. Through in-depth evaluations, we find that low-quality image input is a significant bottleneck in the recognition process. Drawing inspiration from this, we propose the Neighbor-Guided Toolchain Reasoner (NGTR) framework, which is characterized by integrating diverse lightweight tools for visual operations aimed at mitigating issues with low-quality images. Specifically, we transfer a tool selection experience from a similar neighbor to the input and design a reflection module to supervise the tool invocation process. Extensive experiments on public datasets demonstrate that our approach significantly enhances the recognition capabilities of the vanilla VLLMs. We believe that the benchmark and framework could provide an alternative solution to table recognition.

The cameras equipped on autonomous vehicles are frequently contaminated by natural occurrences such as mud spots, stain, and raindrops. Such natural contamination is difficult to detect as abnormality. Inspired by this phenomenon, we propose a novel adversarial camera contamination attack that takes advantage of the inconspicuous nature of lens contamination to achieve object-oriented attacks. Our method specifically targets certain objects (e.g., stop signs) while minimizing the impact on untargeted objects (e.g., cars). By designing a contamination generation module, we model various types of contamination and generate adversarial camera patches that are realistic and effective. Our experiments, conducted in both the digital and physical domains, demonstrate that our method can successfully degrade the detection accuracy of target objects by over 60%, while having minimal impact on untargeted objects.

Cross-domain recommendation (CDR) aims to suggest items from new domains that align with potential user preferences, based on their historical interactions. Existing methods primarily focus on acquiring item representations by discovering user preferences under specific, yet possibly redundant, item features. However, user preferences may be more strongly associated with interacted items at higher semantic levels, rather than specific item features. Consequently, this item feature-focused recommendation approach can easily become suboptimal or even obsolete when conducting CDR with disturbances of these redundant features. In this paper, we propose a novel Preference Prototype-Aware (PPA) learning method to quantitatively learn user preferences while minimizing disturbances from the source domain. The PPA framework consists of two complementary components: a mix-encoder and a preference prototype-aware decoder, forming an end-to-end unified framework suitable for various real-world scenarios. The mix-encoder employs a mix-network to learn better general representations of interacted items and capture the intrinsic relationships between items across different domains. The preference prototype-aware decoder implements a learnable prototype matching mechanism to quantitatively perceive user preferences, which can accurately capture user preferences at a higher semantic level. This decoder can also avoid disturbances caused by item features from the source domain. The experimental results on public benchmark datasets in different scenarios demonstrate the superiority of the proposed PPA learning method compared to state-of-the-art counterparts. PPA excels not only in providing accurate recommendations but also in offering reliable preference prototypes. Our code is available at https://github.com/zyx-nuaa/PPA-for-CDR.

Objective. Developmental Dyslexia (DD) is a special learning disability (SLD), in which it is difficult for children to read, spell and write fluently, seriously affecting children’s related reading skills. Therefore, it is significant for early intervention to quickly and accurately detect children with DD in a no-feeling method. However, most existing detection methods include oral and written assessments, analysis based on brain image (e.g. MRI, fMRI, EEG, etc.), which either require the involvement of domain experts or have low accuracy and are not conducive to large-scale applications. Approach. In this study, inspired by computer vision, we present a novel convolutional-transformer network architecture based on Eye-tracking to improve the diagnosis accuracy of developmental dyslexia in children, which is evaluated on 187 subjects by tracking their eye movements while reading. Main results. To our knowledge, the proposed model achieved a state-of-the-art (SOTA) classification accuracy of 98.21% on the cross-subject dyslexia detection task for the public dataset SDUETDR. Significance. The proposed model can be used in schools as a diagnostic and screening tool for children with developmental dyslexia.

Active Voltage Control (AVC) on the Power Distribution Networks (PDNs) aims to stabilize the voltage levels to ensure efficient and reliable operation of power systems. With the increasing integration of distributed energy resources, recent efforts have explored employing multi-agent reinforcement learning (MARL) techniques to realize effective AVC. Existing methods mainly focus on the acquisition of short-term AVC strategies, i.e., only learning AVC within the short-term training trajectories of a singular diurnal cycle. However, due to the dynamic nature of load demands and renewable energy, the operation states of real-world PDNs may exhibit significant distribution shifts across varying timescales (e.g., daily and seasonal changes). This can render those short-term strategies suboptimal or even obsolete when performing continuous AVC over extended periods. In this paper, we propose a novel temporal prototype-aware learning method, abbreviated as TPA, to learn time-adaptive AVC under short-term training trajectories. At the heart of TPA are two complementary components, namely multi-scale dynamic encoder and temporal prototype-aware policy, that can be readily incorporated into various MARL methods. The former component integrates a stacked transformer network to learn underlying temporal dependencies at different timescales of the PDNs, while the latter implements a learnable prototype matching mechanism to construct a dedicated AVC policy that can dynamically adapt to the evolving operation states. Experimental results on the AVC benchmark with different PDN sizes demonstrate that the proposed TPA surpasses the state-of-the-art counterparts not only in terms of control performance but also by offering model transferability. Our code is available at https://github.com/Canyizl/TPA-for-AVC.

This paper describes our submission to the ICASSP 2023 Signal Processing Grand Challenge (SPGC), which focuses on multilingual Alzheimer’s disease (AD) recognition through spontaneous speech. Our approaches include using a variety of acoustic features and silence-related information for AD detection and mini-mental state examination (MMSE) score prediction, and fine-tuning wav2vec2.0 models on speech in various frequency bands for AD detection. Our overall results on the test data outperform the baseline provided by the organizers, achieving 73.9% accuracy in AD detection by fine-tuning our bilingual wav2vec2.0 pre-trained model on the 0-1000Hz frequency band speech, and 4.610 RMSE (r = 0.565) in MMSE prediction through the fusion of eGeMAPS and silence features.

Alzheimer’s disease is a chronic neurodegenerative disease that usually starts slowly and gradually worsens over time. Although there’s no cure for Alzheimer’s disease yet, a number of recent researches have shown that the early diagnosis and intervention could not only improve the quality of life but also help to slow the progression of the disease. Clock Drawing Test (CDT) is one of the commonly used clinical methods for screening cognitive impairment, due to its simplicity and convenience. In this paper, we’d like to introduce DCDT, a novel Clock Drawing Test system based on digital collection and intellectualized analysis. We first introduce the background of AD and CDT, and then describe the DCDT system from the external and internal aspects. Finally, the demonstration scenario is described briefly.