Recent advancements in integrating Large Language Models (LLM) with automatic speech recognition (ASR) have performed remarkably in general domains. While supervised fine-tuning (SFT) of all model parameters is often employed to adapt pre-trained LLM-based ASR models to specific domains, it imposes high computational costs and notably reduces their performance in general domains. In this paper, we propose a novel parameter-efficient multi-domain fine-tuning method for adapting pre-trained LLM-based ASR models to multi-accent domains without catastrophic forgetting named \textit{HDMoLE}, which leverages hierarchical routing and dynamic thresholds based on combining low-rank adaptation (LoRA) with the mixer of experts (MoE) and can be generalized to any linear layer. Hierarchical routing establishes a clear correspondence between LoRA experts and accent domains, improving cross-domain collaboration among the LoRA experts. Unlike the static Top-K strategy for activating LoRA experts, dynamic thresholds can adaptively activate varying numbers of LoRA experts at each MoE layer. Experiments on the multi-accent and standard Mandarin datasets demonstrate the efficacy of HDMoLE. Applying HDMoLE to an LLM-based ASR model projector module achieves similar performance to full fine-tuning in the target multi-accent domains while using only 9.6% of the trainable parameters required for full fine-tuning and minimal degradation in the source general domain.
相關內容
Recent advances in vision-language foundational models, such as CLIP, have demonstrated significant strides in zero-shot classification. However, the extensive parameterization of models like CLIP necessitates a resource-intensive fine-tuning process. In response, TIP-Adapter and SuS-X have introduced training-free methods aimed at bolstering the efficacy of downstream tasks. While these approaches incorporate support sets to maintain data distribution consistency between knowledge cache and test sets, they often fall short in terms of generalization on the test set, particularly when faced with test data exhibiting substantial distributional variations. In this work, we present CapS-Adapter, an innovative method that employs a caption-based support set, effectively harnessing both image and caption features to exceed existing state-of-the-art techniques in training-free scenarios. CapS-Adapter adeptly constructs support sets that closely mirror target distributions, utilizing instance-level distribution features extracted from multimodal large models. By leveraging CLIP's single and cross-modal strengths, CapS-Adapter enhances predictive accuracy through the use of multimodal support sets. Our method achieves outstanding zero-shot classification results across 19 benchmark datasets, improving accuracy by 2.19\% over the previous leading method. Our contributions are substantiated through extensive validation on multiple benchmark datasets, demonstrating superior performance and robust generalization capabilities. Our code is made publicly available at //github.com/WLuLi/CapS-Adapter.
Network Intrusion Detection Systems (NIDSs) which use machine learning (ML) models achieve high detection performance and accuracy while avoiding dependence on fixed signatures extracted from attack artifacts. However, there is a noticeable hesitance among network security experts and practitioners when it comes to deploying ML-based NIDSs in real-world production environments due to their black-box nature, i.e., how and why the underlying models make their decisions. In this work, we analyze state-of-the-art ML-based online NIDS models using explainable AI (xAI) techniques (e.g., TRUSTEE, SHAP). Using the explanations generated for the models' decisions, the most prominent features used by each NIDS model considered are presented. We compare the explanations generated across xAI methods for a given NIDS model as well as the explanations generated across the NIDS models for a given xAI method. Finally, we evaluate the vulnerability of each NIDS model to inductive bias (artifacts learnt from training data). The results show that: (1) some ML-based NIDS models can be better explained than other models, (2) xAI explanations are in conflict for most of the NIDS models considered in this work and (3) some NIDS models are more vulnerable to inductive bias than other models.
Scientists are increasingly leveraging advances in instruments, automation, and collaborative tools to scale up their experiments and research goals, leading to new bursts of discovery. Various scientific disciplines, including neuroscience, have adopted key technologies to enhance collaboration, reproducibility, and automation. Drawing inspiration from advancements in the software industry, we present a roadmap to enhance the reliability and scalability of scientific operations for diverse research teams tackling large and complex projects. We introduce a five-level Capability Maturity Model describing the principles of rigorous scientific operations in projects ranging from small-scale exploratory studies to large-scale, multi-disciplinary research endeavors. Achieving higher levels of operational maturity necessitates the adoption of new, technology-enabled methodologies, which we refer to as SciOps. This concept is derived from the DevOps methodologies that have revolutionized the software industry. SciOps involves digital research environments that seamlessly integrate computational, automation, and AI-driven efforts throughout the research cycle-from experimental design and data collection to analysis and dissemination, ultimately leading to closed-loop discovery. This maturity model offers a framework for assessing and improving operational practices in multidisciplinary research teams, guiding them towards greater efficiency and effectiveness in scientific inquiry.
Recent advancements in 3D Gaussian Splatting (3DGS) have unlocked significant potential for modeling 3D head avatars, providing greater flexibility than mesh-based methods and more efficient rendering compared to NeRF-based approaches. Despite these advancements, the creation of controllable 3DGS-based head avatars remains time-intensive, often requiring tens of minutes to hours. To expedite this process, we here introduce the "Gaussian Deja-vu" framework, which first obtains a generalized model of the head avatar and then personalizes the result. The generalized model is trained on large 2D (synthetic and real) image datasets. This model provides a well-initialized 3D Gaussian head that is further refined using a monocular video to achieve the personalized head avatar. For personalizing, we propose learnable expression-aware rectification blendmaps to correct the initial 3D Gaussians, ensuring rapid convergence without the reliance on neural networks. Experiments demonstrate that the proposed method meets its objectives. It outperforms state-of-the-art 3D Gaussian head avatars in terms of photorealistic quality as well as reduces training time consumption to at least a quarter of the existing methods, producing the avatar in minutes.
By leveraging both texts and images, large vision language models (LVLMs) have shown significant progress in various multi-modal tasks. Nevertheless, these models often suffer from hallucinations, e.g., they exhibit inconsistencies between the visual input and the textual output. To address this, we propose H-POPE, a coarse-to-fine-grained benchmark that systematically assesses hallucination in object existence and attributes. Our evaluation shows that models are prone to hallucinations on object existence, and even more so on fine-grained attributes. We further investigate whether these models rely on visual input to formulate the output texts.
Recent advancements in Intrusion Detection Systems (IDS), integrating Explainable AI (XAI) methodologies, have led to notable improvements in system performance via precise feature selection. However, a thorough understanding of cyber-attacks requires inherently explainable decision-making processes within IDS. In this paper, we present the Interpretable Generalization Mechanism (IG), poised to revolutionize IDS capabilities. IG discerns coherent patterns, making it interpretable in distinguishing between normal and anomalous network traffic. Further, the synthesis of coherent patterns sheds light on intricate intrusion pathways, providing essential insights for cybersecurity forensics. By experiments with real-world datasets NSL-KDD, UNSW-NB15, and UKM-IDS20, IG is accurate even at a low ratio of training-to-test. With 10%-to-90%, IG achieves Precision (PRE)=0.93, Recall (REC)=0.94, and Area Under Curve (AUC)=0.94 in NSL-KDD; PRE=0.98, REC=0.99, and AUC=0.99 in UNSW-NB15; and PRE=0.98, REC=0.98, and AUC=0.99 in UKM-IDS20. Notably, in UNSW-NB15, IG achieves REC=1.0 and at least PRE=0.98 since 40%-to-60%; in UKM-IDS20, IG achieves REC=1.0 and at least PRE=0.88 since 20%-to-80%. Importantly, in UKM-IDS20, IG successfully identifies all three anomalous instances without prior exposure, demonstrating its generalization capabilities. These results and inferences are reproducible. In sum, IG showcases superior generalization by consistently performing well across diverse datasets and training-to-test ratios (from 10%-to-90% to 90%-to-10%), and excels in identifying novel anomalies without prior exposure. Its interpretability is enhanced by coherent evidence that accurately distinguishes both normal and anomalous activities, significantly improving detection accuracy and reducing false alarms, thereby strengthening IDS reliability and trustworthiness.
Parameter-Efficient Fine-Tuning (PEFT) methods have gained significant popularity for adapting pre-trained Large Language Models (LLMs) to downstream tasks, primarily due to their potential to significantly reduce memory and computational overheads. However, a common limitation in most PEFT approaches is their application of a uniform architectural design across all layers. This uniformity involves identical trainable modules and ignores the varying importance of each layer, leading to sub-optimal fine-tuning results. To overcome the above limitation and obtain better performance, we develop a novel approach, Importance-aware Sparse Tuning (IST), to fully utilize the inherent sparsity and select the most important subset of full layers with effective layer-wise importance scoring. The proposed IST is a versatile and plug-and-play technique compatible with various PEFT methods that operate on a per-layer basis. By leveraging the estimated importance scores, IST dynamically updates these selected layers in PEFT modules, leading to reduced memory demands. We further provide theoretical proof of convergence and empirical evidence of superior performance to demonstrate the advantages of IST over uniform updating strategies. Extensive experiments on a range of LLMs, PEFTs, and downstream tasks substantiate the effectiveness of our proposed method, showcasing IST's capacity to enhance existing layer-based PEFT methods. Our code is available at //github.com/Kaiseem/IST.
Recent advancements in automatic code generation using large language models (LLMs) have brought us closer to fully automated secure software development. However, existing approaches often rely on a single agent for code generation, which struggles to produce secure, vulnerability-free code. Traditional program synthesis with LLMs has primarily focused on functional correctness, often neglecting critical dynamic security implications that happen during runtime. To address these challenges, we propose AutoSafeCoder, a multi-agent framework that leverages LLM-driven agents for code generation, vulnerability analysis, and security enhancement through continuous collaboration. The framework consists of three agents: a Coding Agent responsible for code generation, a Static Analyzer Agent identifying vulnerabilities, and a Fuzzing Agent performing dynamic testing using a mutation-based fuzzing approach to detect runtime errors. Our contribution focuses on ensuring the safety of multi-agent code generation by integrating dynamic and static testing in an iterative process during code generation by LLM that improves security. Experiments using the SecurityEval dataset demonstrate a 13% reduction in code vulnerabilities compared to baseline LLMs, with no compromise in functionality.
Recent advancements in Augmented Reality (AR) research have highlighted the critical role of context awareness in enhancing interface effectiveness and user experience. This underscores the need for intelligent AR (iAR) interfaces that dynamically adapt across various contexts to provide optimal experiences. In this paper, we (a) propose a comprehensive framework for context-aware inference and adaptation in iAR, (b) introduce a taxonomy that describes context through quantifiable input data, and (c) present an architecture that outlines the implementation of our proposed framework and taxonomy within iAR. Additionally, we present an empirical AR experiment to observe user behavior and record user performance, context, and user-specified adaptations to the AR interfaces within a context-switching scenario. We (d) explore the nuanced relationships between context and user adaptations in this scenario and discuss the significance of our framework in identifying these patterns. This experiment emphasizes the significance of context-awareness in iAR and provides a preliminary training dataset for this specific Scenario.
Object detection with transformers (DETR) reaches competitive performance with Faster R-CNN via a transformer encoder-decoder architecture. Inspired by the great success of pre-training transformers in natural language processing, we propose a pretext task named random query patch detection to unsupervisedly pre-train DETR (UP-DETR) for object detection. Specifically, we randomly crop patches from the given image and then feed them as queries to the decoder. The model is pre-trained to detect these query patches from the original image. During the pre-training, we address two critical issues: multi-task learning and multi-query localization. (1) To trade-off multi-task learning of classification and localization in the pretext task, we freeze the CNN backbone and propose a patch feature reconstruction branch which is jointly optimized with patch detection. (2) To perform multi-query localization, we introduce UP-DETR from single-query patch and extend it to multi-query patches with object query shuffle and attention mask. In our experiments, UP-DETR significantly boosts the performance of DETR with faster convergence and higher precision on PASCAL VOC and COCO datasets. The code will be available soon.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191