AAAI 2025 Papers

As multimedia information proliferates, multimodal recommendation systems have garnered significant attention. These systems leverage multimodal information to alleviate the data sparsity issue inherent in recommendation systems, thereby enhancing the accuracy of recommendations. Due to the natural semantic disparities among multimodal features, recent research has primarily focused on cross-modal alignment using self-supervised learning to bridge these gaps. However, aligning different modal features might result in the loss of valuable interaction information, distancing them from ID embeddings. It is crucial to recognize that the primary goal of multimodal recommendation is to predict user preferences, not merely to understand multimodal content. To this end, we propose a new Multi-level sElf-supervised learNing for mulTimOdal Recommendation (MENTOR) method, which effectively reduces the gap among modalities while retaining interaction information. Specifically, MENTOR begins by extracting representations from each modality using both heterogeneous user-item and homogeneous item-item graphs. It then employs a multilevel cross-modal alignment task, guided by ID embeddings, to align modalities across multiple levels while retaining historical interaction information. To balance effectiveness and efficiency, we further propose an optional general feature enhancement task that bolsters the general features from both structure and feature perspectives, thus enhancing the robustness of our model.

Automatic Speech Recognition (ASR) systems are pivotal in transcribing speech into text, yet the errors they introduce can significantly degrade the performance of downstream tasks like summarization. This issue is particularly pronounced in clinical dialogue summarization, a low-resource domain where supervised data for fine-tuning is scarce, necessitating the use of ASR models as black-box solutions. Employing conventional data augmentation for enhancing the noise robustness of summarization models is not feasible either due to the unavailability of sufficient medical dialogue audio recordings and corresponding ASR transcripts. To address this challenge, we propose MEDSAGE, an approach for generating synthetic samples for data augmentation using Large Language Models (LLMs). Specifically, we leverage the in-context learning capabilities of LLMs and instruct them to generate ASR-like errors based on a few available medical dialogue examples with audio recordings. Experimental results show that LLMs can effectively model ASR noise, and incorporating this noisy data into the training process significantly improves the robustness and accuracy of medical dialogue summarization systems. This approach addresses the challenges of noisy ASR outputs in critical applications, offering a robust solution to enhance the reliability of clinical dialogue summarization.

The detection of anomalous tissue regions (ATRs) within affected tissues is crucial in clinical diagnosis and pathological studies. Conventional automated ATR detection methods, primarily based on histology images alone, falter in cases where ATRs and normal tissues have subtle visual differences. The recent spatial transcriptomics (ST) technology profiles gene expressions across tissue regions, offering a molecular perspective for detecting ATRs. However, there is a dearth of ATR detection methods that effectively harness complementary information from both histology images and ST. To address this gap, we propose MEATRD, a novel ATR detection method that integrates histology image and ST data. MEATRD is trained to reconstruct image patches and gene expression profiles of normal tissue spots (inliers) from their multimodal embeddings, followed by learning a one-class classification AD model based on latent multimodal reconstruction errors. This strategy harmonizes the strengths of reconstruction-based and one-class classification approaches. At the heart of MEATRD is an innovative masked graph dual-attention transformer (MGDAT) network, which not only facilitates cross-modality and cross-node information sharing but also addresses the model over-generalization issue commonly seen in reconstruction-based AD methods. Additionally, we demonstrate that modality-specific, task-relevant information is collated and condensed in multimodal bottleneck encoding generated in MGDAT, marking the first theoretical analysis of the informational properties of multimodal bottleneck encoding. Extensive evaluations across eight real ST datasets reveal MEATRD's superior performance in ATR detection, surpassing various state-of-the-art AD methods. Remarkably, MEATRD also proves adept at discerning ATRs that only show slight visual deviations from normal tissues.

Value alignment, at the intersection of moral philosophy and AI safety, is dedicated to ensuring that artificially intelligent (AI) systems align with a certain set of values. One challenge facing value alignment researchers is accurately translating these values into a machine readable format. In the case of reinforcement learning (RL), a popular method within value alignment, this requires designing a reward function which accurately defines the value of all state-action pairs. It is common for programmers to hand-set and manually tune these values. In this paper, we examine the challenges of hand-programming values into reward functions for value alignment, and propose mathematical models as an alternative grounding for reward function design in ethical scenarios. Experimental results demonstrate that our modelled-ethics approach offers a more consistent alternative and outperforms our hand-programmed reward functions.

Fatigue is a critical factor contributing to accidents in industries such as safety monitoring and engineering construction. Fatigue exhibits dynamic complexity and non-stationary characteristics, so there are many intermediate states of short-term variation between alert and fatigue. Capturing and learning the signs of these intermediate states is essential for accurate fatigue assessment. However, current fatigue detection methods primarily rely on coarse-grained labels, typically spanning minutes to hours, and commonly treat alert and fatigue as two distinctly separate distributions, overlooking the expression of intermediate states and oversimplifying the rich distribution information of fatigue types and levels, thereby limiting detection effectiveness. To address these, this paper explores a refined representation of fatigue in terms of three dimensions: time, type, and level, and proposes a Multi-Dimensional Fine-Grained Modeling for Fatigue Detection (MDFG). This introduces the SmallLoss to extract trustworthy samples, utilizes clustering to identify diverse subtypes under alert and fatigued states, and establishes base class sets in each state. Subsequently, a complete base class set containing intermediate state bases is constructed using the base class synthesis method, which achieves the expression of intermediate fatigue states from absence to presence. Finally, fatigue levels are quantified based on the matching between samples and the complete base class set. Moreover, to cope with the complex variability of fatigue states, MDFG employs meta-learning for training. MDFG achieves an Average accuracy improvement of 10.0% and 12.1% on two real datasets compared to methods that do not consider fine-grained information. Extensive experiments demonstrate that the MDFG exhibits superior robustness and stability among current fatigue detection methods.

The clinical diagnosis of most mental disorders primarily relies on the conversations between psychiatrist and patient. The creation of such diagnostic conversation datasets is promising to boost the AI mental healthcare community. However, directly collecting the conversations in real diagnosis scenarios is near impossible due to stringent privacy and ethical considerations. To address this issue, we seek to synthesize diagnostic conversation by exploiting anonymized patient cases that are easier to access. Specifically, we design a neuro-symbolic multi-agent framework for synthesizing the diagnostic conversation of mental disorders with large language models. It takes patient case as input and is capable of generating multiple diverse conversations with one single patient case. The framework basically involves the interaction between a doctor agent and a patient agent, and generates conversations under symbolic control via a dynamic diagnosis tree. By applying the proposed framework, we develop the largest Chinese mental disorders diagnosis dataset MDD-5k. This dataset is built upon 1000 real, anonymized patient cases by cooperating with Shanghai Mental Health Center and comprises 5000 high-quality long conversations with diagnosis results and treatment opinions as labels. To the best of our knowledge, it's also the first labeled dataset for Chinese mental disorders diagnosis. Human evaluation demonstrates the proposed MDD-5k dataset successfully simulates human-like diagnostic process of mental disorders.

Generative adversarial networks (GANs) have emerged as a powerful tool for generating high-fidelity data. However, the main bottleneck of existing approaches is the lack of supervision on the generator training, which often results in undamped oscillation and unsatisfactory performance. To address this issue, we propose an algorithm called Monte Carlo GAN (MCGAN). This approach, utilizing an innovative generative loss function, termed the regression loss, reformulates the generator training as a regression task and enables the generator training by minimizing the mean squared error between the discriminator's output of real data and the expected discriminator of fake data. We demonstrate the desirable analytic properties of the regression loss, including discriminability and optimality, and show that our method requires a weaker condition on the discriminator for effective generator training. These properties justify the strength of this approach to improve the training stability while retaining the optimality of GAN by leveraging strong supervision of the regression loss. Extensive experiments on diverse datasets, including image data (CIFAR-10/100, FFHQ256, ImageNet, and LSUN Bedroom), time series data (VAR and stock data), and video data, are conducted to demonstrate the flexibility and effectiveness of our proposed MCGAN. Numerical results show that the proposed MCGAN is versatile in enhancing a variety of backbone GAN models and achieves consistent and significant improvement in terms of quality, accuracy, training stability, and learned latent space.

Accurate standard plane acquisition in fetal ultrasound (US) videos is crucial for fetal growth assessment, anomaly detection, and adherence to clinical guidelines. However, manually selecting standard frames is time-consuming and prone to intra- and inter-sonographer variability. Existing methods primarily rely on image-based approaches that capture standard frames and then classify the input frames across different anatomies. This ignores the dynamic nature of video acquisition and its interpretation. To address these challenges, we introduce Multi-Tier Class-Aware Token Transformer (MCAT); a visual query-based video clip localization (VQ-VCL) method to assist sonographers by enabling them to capture a quick US sweep. By then providing a visual query of the anatomy they wish to analyze, MCAT returns the video clip containing the standard frames for that anatomy, facilitating thorough screening for potential anomalies. We evaluate MCAT on two ultrasound video datasets and a natural image VQ-VCL dataset based on Ego4D. Our model outperforms state-of-the-art methods by 10% and 13% mtIoU on the ultrasound datasets and by 5.35% mtIoU on the Ego4D dataset, using 96% fewer tokens. MCAT’s efficiency and accuracy have significant potential implications for public health, especially in low- and middle-income countries (LMICs), where it may enhance prenatal care by streamlining standard plane acquisition, simplifying US based screening, diagnosis and allowing sonographers to examine more patients.

Matching markets, in which agents are assigned to one another based on preferences and capacity constraints, are pervasive in various domains. This paper introduces MATWA (https://matwa.optimalmatching.com), a web application that offers the most comprehensive collection to date of algorithms for fundamental matching under preference problem classes. MATWA provides results of algorithm executions and visualisations of structural properties. It is intended to be a resource for the community of researchers, educators and practitioners, supporting experimentation, as well as aiding the understanding of matching algorithms.

Pretrained visual-language models have made significant advancements in multimodal tasks, including image-text retrieval. However, a major challenge in image-text matching lies in language bias, where models predominantly rely on language priors and neglect to adequately consider the visual content. We thus present Multimodal ASsociation Score (MASS), a framework that reduces the reliance on language priors for better visual accuracy in image-text matching problems. It can be seamlessly incorporated into existing visual-language models without necessitating additional training. Our experiments have shown that \modelname effectively lessens language bias without losing an understanding of linguistic compositionality. Overall, MASS offers a promising solution for enhancing image-text matching performance in visual-language models.

Auto-regressive models have made significant progress in the realm of text-to-image synthesis, yet devising an appropriate model architecture and training strategy to achieve a satisfactory level remains an important avenue of exploration. In this work, we introduce MARS, a novel framework for T2I generation that incorporates a specially designed Semantic Vision-Language Integration Expert (SemVIE). This innovative component integrates pre-trained LLMs by independently processing linguistic and visual information—freezing the textual component while fine-tuning the visual component. This methodology preserves the NLP capabilities of LLMs while imbuing them with exceptional visual understanding. Building upon the powerful base of the pre-trained Qwen-7B, MARS stands out with its bilingual generative capabilities corresponding to both English and Chinese language prompts and the capacity for joint image and text generation. The flexibility of this framework lends itself to migration towards any-to-any task adaptability. Furthermore, MARS employs a multi-stage training strategy that first establishes robust image-text alignment through complementary bidirectional tasks and subsequently concentrates on refining the T2I generation process, significantly augmenting text-image synchrony and the granularity of image details. Notably, MARS requires only 9% of the GPU days needed by SD1.5, yet it achieves remarkable results across a variety of benchmarks, illustrating the training efficiency and the potential for swift deployment in various applications.

The advent of large language models (LLMs) has sparked significant interest in using natural language for preference learning. However, existing methods often suffer from high computational burdens, taxing human supervision, and lack of interpretability. To address these issues, we introduce MAPLE, a framework for large language model-guided Bayesian active preference learning. MAPLE leverages LLMs to model the distribution over preference functions, conditioning it on both natural language feedback and conventional preference learning feedback, such as pairwise trajectory rankings. MAPLE employs active learning to systematically reduce uncertainty in this distribution and incorporates a language-conditioned active query selection mechanism to identify informative and easy-to-answer queries, thus reducing the burden on humans. We evaluate MAPLE's sample efficiency and preference inference quality across two benchmarks, including a real-world vehicle route planning benchmark using OpenStreetMap data. Our results demonstrate that MAPLE accelerates the learning process and effectively improves humans' ability to answer queries.

Multi-agent pathfinding (MAPF) is a problem that generally requires finding collision-free paths for multiple agents in a shared environment. Solving MAPF optimally, even under restrictive assumptions, is NP-hard, yet efficient solutions for this problem are critical for numerous applications, such as automated warehouses and transportation systems. Recently, learning-based approaches to MAPF have gained attention, particularly those leveraging deep reinforcement learning. Typically, such learning-based MAPF solvers are augmented with additional components like single-agent planning or communication. Orthogonally, in this work we rely solely on imitation learning that leverages a large dataset of expert MAPF solutions and transformer-based neural network to create a foundation model for MAPF called MAPF-GPT. The latter is capable of generating actions without additional heuristics or communication. MAPF-GPT demonstrates zero-shot learning abilities when solving the MAPF problems that are not present in the training dataset. We show that MAPF-GPT notably outperforms the current best-performing learnable MAPF solvers on a diverse range of problem instances and is computationally efficient during inference.

Multi-modal transformers are rapidly gaining attention in video captioning tasks. Existing multi-modal video captioning methods extract a fixed number of frames, but this has critical challenges. If a limited number of frames are extracted, important frames with essential information for caption generation may be missed. Conversely, extracting an excessive number of frames includes consecutive frames, potentially causing redundancy in visual tokens extracted from consecutive video frames. To extract an appropriate number of frames for each video, this paper proposes the first model-agnostic module selection framework in video captioning that has two main functions: (1) selecting a caption generation module with an appropriate size based on visual tokens extracted from video frames, and (2) constructing subsets of visual tokens for the selected caption generation module. Furthermore, we propose a new adaptive attention masking scheme that enhances attention on important visual tokens. Our numerical experiments with three different benchmark datasets demonstrate that the proposed framework significantly improves the performances of three recent video captioning models.

Self-correction in text-to-SQL is the process of prompting large language model (LLM) to revise its previously incorrectly generated SQL, and commonly relies on manually crafted self-correction guidelines by human experts that are not only labor-intensive to produce but also limited by the human ability in identifying all potential error patterns in LLM responses. We introduce MAGIC, a novel multi-agent method that automates the creation of the self-correction guideline. MAGIC uses three specialized agents: a manager, a correction, and a feedback agent. These agents collaborate on the failures of an LLM-based method on the training set to iteratively generate and refine a self-correction guideline tailored to LLM mistakes, mirroring human processes but without human involvement. Our extensive experiments show that MAGIC's guideline outperforms expert human's created ones. We empirically find out that the guideline produced by MAGIC enhances the interpretability of the corrections made, providing insights in analyzing the reason behind the failures and successes of LLMs in self-correction.

This paper proposes MAFT, a novel multimodal automated fact-checking system capable of handling content in any combination of text, images, videos, and audio. The core idea behind our system is the textualization of multimodal content using various machine learning techniques. MAFT comprehensively analyzes this textualized content along with external information collected via web APIs by large language models (LLMs). MAFT generates interpretable fact-checking reports that include not only verification results but also a detailed verification process. With its adaptability and ability to automatically verify multimodal content, MAFT contributes to the fight against the spread of multimodal misinformation.

Models trained on real-world data often mirror and exacerbate existing social biases. Traditional methods for mitigating these biases typically require prior knowledge of the specific biases to be addressed, and the social groups associated with each instance. In this paper, we introduce a novel adversarial training strategy that operates withour relying on prior bias-type knowledge (e.g., gender or racial bias) and protected attribute labels. Our approach dynamically identifies biases during model training by utilizing auxiliary bias detector. These detected biases are simultaneously mitigated through adversarial training. Crucially, we implement these bias detectors at various levels of the feature maps of the main model, enabling the detection of a broader and more nuanced range of bias features. Through experiments on racial and gender biases in sentiment and occupation classification tasks, our method effectively reduces social biases without the need for demographic annotations. Moreover, our approach not only matches but often surpasses the efficacy of methods that require detailed demographic insights, marking a significant advancement in bias mitigation techniques.

The perception system for autonomous driving generally requires to handle multiple diverse sub-tasks. However, current algorithms typically tackle individual sub-tasks separately, which leads to low efficiency when aiming at obtaining full-perception results. Some multi-task learning methods try to unify multiple tasks with one model, but do not solve the conflicts in multi-task learning. In this paper, we introduce M3Net, a novel multimodal and multi-task network that simultaneously tackles detection, segmentation, and 3D occupancy prediction for autonomous driving and achieves superior performance than single task model. M3Net takes multimodal data as input and multiple tasks via query-token interactions. To enhance the integration of multi-modal features for multi-task learning, we first propose the Modality-Adaptive Feature Integration (MAFI) module, which enables single-modality features to predict channel-wise attention weights for their high-performing tasks, respectively. Based on integrated features, we then develop task-specific query initialization strategies to accommodate the needs of detection/segmentation and 3D occupancy prediction. Leveraging the properly initialized queries, a shared decoder transforms queries and BEV features layer-wise, facilitating multi-task learning. Furthermore, we propose a Task-oriented Channel Scaling (TCS) module in the decoder to mitigate conflicts between optimizing for different tasks. Additionally, our proposed multi-task querying and TCS module support both Transformer-based decoder and Mamba-based decoder, demonstrating its flexibility to different architectures. M3Net achieves state-of-the-art multi-task learning performance on the nuScenes benchmarks.

Audio classification plays a crucial role within fields such as human-machine interaction and intelligent robotics. However, high-performance audio classification systems typically demand significant computational and storage resources, posing substantial challenges when deploying to the resource-constrained edge devices with an urgent need for such capabilities. To achieve a new level of balance between model complexity and performance, we introduce a novel multi-view method for the separated time-frequency features extraction and utilization, which exists within the proposed Mini Mirror Multi-View Network (M3Net) in the form of the Mirror Attention mechanism. M3Net enables reversible spatial transformation of spectral features is capable of efficiently leverages robust local and global features in the time and frequency domains with low requirements for parameters. Experiments based on Mel-Spectrogram without data augmentation and pre-training indicate that M3Net can achieve classification accuracy over 97% on the UrbanSound8K and SpeechCommandsV2 datasets with only 0.03 million parameters. The contribution of each functional segment in M3Net is fully verified and explained in the ablation experiments.