Abstract:Graph classification is a critical task in numerous multimedia applications, where graphs are employed to represent diverse types of multimedia data, including images, videos, and social networks. Nevertheless, in real-world scenarios, labeled graph data can be limited or scarce. To address this issue, we focus on the problem of semi-supervised graph classification, which involves both supervised and unsupervised models learning from labeled and unlabeled data. In contrast to recent approaches that transfer the entire knowledge from the unsupervised model to the supervised one, we argue that an effective transfer should only retain the relevant semantics that align well with the supervised task. In this paper, we propose a novel framework named DisenSemi, which learns disentangled representation for semi-supervised graph classification. Specifically, a disentangled graph encoder is proposed to generate factor-wise graph representations for both supervised and unsupervised models. Then we train two models via supervised objective and mutual information (MI)-based constraints respectively. To ensure the meaningful transfer of knowledge from the unsupervised encoder to the supervised one, we further define an MI-based disentangled consistency regularization between two models and identify the corresponding rationale that aligns well with the current graph classification task. Experimental results on a range of publicly accessible datasets reveal the effectiveness of our DisenSemi.
Abstract:Recently, with the rise of web videos, managing and understanding large-scale video datasets has become increasingly important. Video Large Language Models (VideoLLMs) have emerged in recent years due to their strong video understanding capabilities. However, training and inference processes for VideoLLMs demand vast amounts of data, presenting significant challenges to data management, particularly regarding efficiency, robustness, and effectiveness. In this work, we present KeyVideoLLM, a text-video frame similarity-based keyframe selection method designed to manage VideoLLM data efficiently, robustly, and effectively. Specifically, KeyVideoLLM achieves a remarkable data compression rate of up to 60.9 times, substantially lowering disk space requirements, which proves its high efficiency. Additionally, it maintains a 100% selection success rate across all video formats and scales, enhances processing speed by up to 200 times compared to existing keyframe selection methods, and does not require hyperparameter tuning. Beyond its outstanding efficiency and robustness, KeyVideoLLM further improves model performance in video question-answering tasks during both training and inference stages. Notably, it consistently achieved the state-of-the-art (SoTA) experimental results on diverse datasets.
Abstract:Adapting Large Language Models (LLMs) for recommendation requires careful consideration of the decoding process, given the inherent differences between generating items and natural language. Existing approaches often directly apply LLMs' original decoding methods. However, we find these methods encounter significant challenges: 1) amplification bias -- where standard length normalization inflates scores for items containing tokens with generation probabilities close to 1 (termed ghost tokens), and 2) homogeneity issue -- generating multiple similar or repetitive items for a user. To tackle these challenges, we introduce a new decoding approach named Debiasing-Diversifying Decoding (D3). D3 disables length normalization for ghost tokens to alleviate amplification bias, and it incorporates a text-free assistant model to encourage tokens less frequently generated by LLMs for counteracting recommendation homogeneity. Extensive experiments on real-world datasets demonstrate the method's effectiveness in enhancing accuracy and diversity.
Abstract:Nested Named Entity Recognition (NNER) focuses on addressing overlapped entity recognition. Compared to Flat Named Entity Recognition (FNER), annotated resources are scarce in the corpus for NNER. Data augmentation is an effective approach to address the insufficient annotated corpus. However, there is a significant lack of exploration in data augmentation methods for NNER. Due to the presence of nested entities in NNER, existing data augmentation methods cannot be directly applied to NNER tasks. Therefore, in this work, we focus on data augmentation for NNER and resort to more expressive structures, Composited-Nested-Label Classification (CNLC) in which constituents are combined by nested-word and nested-label, to model nested entities. The dataset is augmented using the Composited-Nested-Learning (CNL). In addition, we propose the Confidence Filtering Mechanism (CFM) for a more efficient selection of generated data. Experimental results demonstrate that this approach results in improvements in ACE2004 and ACE2005 and alleviates the impact of sample imbalance.
Abstract:Background: Accelerated real-time cine (RT-Cine) imaging enables cardiac function assessment without the need for breath-holding. However, when performed during in-magnet exercise, RT-Cine images may exhibit significant motion artifacts. Methods: By projecting the time-averaged images to the subspace spanned by the coil sensitivity maps, we propose a coil reweighting (CR) method to automatically suppress a subset of receive coils that introduces a high level of artifacts in the reconstructed image. RT-Cine data collected at rest and during exercise from ten healthy volunteers and six patients were utilized to assess the performance of the proposed method. One short-axis and one two-chamber RT-Cine series reconstructed with and without CR from each subject were visually scored by two cardiologists in terms of the level of artifacts on a scale of 1 (worst) to 5 (best). Results: For healthy volunteers, applying CR to RT-Cine images collected at rest did not significantly change the image quality score (p=1). In contrast, for RT-Cine images collected during exercise, CR significantly improved the score from 3.9 to 4.68 (p<0.001). Similarly, in patients, CR did not significantly change the score for images collected at rest (p=0.031) but markedly improved the score from 3.15 to 4.42 (p<0.001) for images taken during exercise. Despite lower image quality scores in the patient cohort compared to healthy subjects, likely due to larger body habitus and the difficulty of limiting body motion during exercise, CR effectively suppressed motion artifacts, with all image series from the patient cohort receiving a score of four or higher. Conclusion: Using data from healthy subjects and patients, we demonstrate that the motion artifacts in the reconstructed RT-Cine images can be effectively suppressed significantly with the proposed CR method.
Abstract:Image-text matching has been a long-standing problem, which seeks to connect vision and language through semantic understanding. Due to the capability to manage large-scale raw data, unsupervised hashing-based approaches have gained prominence recently. They typically construct a semantic similarity structure using the natural distance, which subsequently provides guidance to the model optimization process. However, the similarity structure could be biased at the boundaries of semantic distributions, causing error accumulation during sequential optimization. To tackle this, we introduce a novel hashing approach termed Distribution-based Structure Mining with Consistency Learning (DEMO) for efficient image-text matching. From a statistical view, DEMO characterizes each image using multiple augmented views, which are considered as samples drawn from its intrinsic semantic distribution. Then, we employ a non-parametric distribution divergence to ensure a robust and precise similarity structure. In addition, we introduce collaborative consistency learning which not only preserves the similarity structure in the Hamming space but also encourages consistency between retrieval distribution from different directions in a self-supervised manner. Through extensive experiments on three benchmark image-text matching datasets, we demonstrate that DEMO achieves superior performance compared with many state-of-the-art methods.
Abstract:Large Language Models (LLMs) like ChatGPT and GPT-4 are versatile and capable of addressing a diverse range of tasks. However, general LLMs, which are developed on open-domain data, may lack the domain-specific knowledge essential for tasks in vertical domains, such as legal, medical, etc. To address this issue, previous approaches either conduct continuous pre-training with domain-specific data or employ retrieval augmentation to support general LLMs. Unfortunately, these strategies are either cost-intensive or unreliable in practical applications. To this end, we present a novel framework named BLADE, which enhances Black-box LArge language models with small Domain-spEcific models. BLADE consists of a black-box LLM and a small domain-specific LM. The small LM preserves domain-specific knowledge and offers specialized insights, while the general LLM contributes robust language comprehension and reasoning capabilities. Specifically, our method involves three steps: 1) pre-training the small LM with domain-specific data, 2) fine-tuning this model using knowledge instruction data, and 3) joint Bayesian optimization of the general LLM and the small LM. Extensive experiments conducted on public legal and medical benchmarks reveal that BLADE significantly outperforms existing approaches. This shows the potential of BLADE as an effective and cost-efficient solution in adapting general LLMs for vertical domains.
Abstract:Recent research demonstrates the effectiveness of using pre-trained language models for legal case retrieval. Most of the existing works focus on improving the representation ability for the contextualized embedding of the [CLS] token and calculate relevance using textual semantic similarity. However, in the legal domain, textual semantic similarity does not always imply that the cases are relevant enough. Instead, relevance in legal cases primarily depends on the similarity of key facts that impact the final judgment. Without proper treatments, the discriminative ability of learned representations could be limited since legal cases are lengthy and contain numerous non-key facts. To this end, we introduce DELTA, a discriminative model designed for legal case retrieval. The basic idea involves pinpointing key facts in legal cases and pulling the contextualized embedding of the [CLS] token closer to the key facts while pushing away from the non-key facts, which can warm up the case embedding space in an unsupervised manner. To be specific, this study brings the word alignment mechanism to the contextual masked auto-encoder. First, we leverage shallow decoders to create information bottlenecks, aiming to enhance the representation ability. Second, we employ the deep decoder to enable translation between different structures, with the goal of pinpointing key facts to enhance discriminative ability. Comprehensive experiments conducted on publicly available legal benchmarks show that our approach can outperform existing state-of-the-art methods in legal case retrieval. It provides a new perspective on the in-depth understanding and processing of legal case documents.
Abstract:Collecting relevant judgments for legal case retrieval is a challenging and time-consuming task. Accurately judging the relevance between two legal cases requires a considerable effort to read the lengthy text and a high level of domain expertise to extract Legal Facts and make juridical judgments. With the advent of advanced large language models, some recent studies have suggested that it is promising to use LLMs for relevance judgment. Nonetheless, the method of employing a general large language model for reliable relevance judgments in legal case retrieval is yet to be thoroughly explored. To fill this research gap, we devise a novel few-shot workflow tailored to the relevant judgment of legal cases. The proposed workflow breaks down the annotation process into a series of stages, imitating the process employed by human annotators and enabling a flexible integration of expert reasoning to enhance the accuracy of relevance judgments. By comparing the relevance judgments of LLMs and human experts, we empirically show that we can obtain reliable relevance judgments with the proposed workflow. Furthermore, we demonstrate the capacity to augment existing legal case retrieval models through the synthesis of data generated by the large language model.
Abstract:Background: Cardiovascular magnetic resonance imaging (CMR) is a well-established imaging tool for diagnosing and managing cardiac conditions. The integration of exercise stress with CMR (ExCMR) can enhance its diagnostic capacity. Despite recent advances in CMR technology, ExCMR remains technically challenging due to motion artifacts and limited spatial and temporal resolution. Methods: This study investigates the feasibility of biventricular functional and hemodynamic assessment using real-time (RT) ExCMR during a staged exercise protocol in 26 healthy volunteers. We introduce a coil reweighting technique to minimize motion artifacts. In addition, we identify and analyze heartbeats from the end-expiratory phase to enhance the repeatability of cardiac function quantification. To demonstrate clinical feasibility, qualitative results from five patients are also presented. Results: Our findings indicate a consistent decrease in end-systolic volume (ESV) and stable end-diastolic volume (EDV) across exercise intensities, leading to increased stroke volume (SV) and ejection fraction (EF). Coil reweighting effectively reduces motion artifacts, improving image quality in both healthy volunteers and patients. The repeatability of cardiac function parameters, demonstrated by scan-rescan tests in nine volunteers, improves with the selection of end-expiratory beats. Conclusions: The study demonstrates that RT ExCMR with in-magnet exercise is a feasible and effective method for dynamic cardiac function monitoring during exercise. The proposed coil reweighting technique and selection of end-expiratory beats significantly enhance image quality and repeatability.