This paper proposes an efficient attempt to noisy speech emotion recognition (NSER). Conventional NSER approaches have proven effective in mitigating the impact of artificial noise sources, such as white Gaussian noise, but are limited to non-stationary noises in real-world environments due to their complexity and uncertainty. To overcome this limitation, we introduce a new method for NSER by adopting the automatic speech recognition (ASR) model as a noise-robust feature extractor to eliminate non-vocal information in noisy speech. We first obtain intermediate layer information from the ASR model as a feature representation for emotional speech and then apply this representation for the downstream NSER task. Our experimental results show that 1) the proposed method achieves better NSER performance compared with the conventional noise reduction method, 2) outperforms self-supervised learning approaches, and 3) even outperforms text-based approaches using ASR transcription or the ground truth transcription of noisy speech.
相關內容
This paper explores the frontiers of large language models (LLMs) in psychology applications. Psychology has undergone several theoretical changes, and the current use of Artificial Intelligence (AI) and Machine Learning, particularly LLMs, promises to open up new research directions. We provide a detailed exploration of how LLMs like ChatGPT are transforming psychological research. It discusses the impact of LLMs across various branches of psychology, including cognitive and behavioral, clinical and counseling, educational and developmental, and social and cultural psychology, highlighting their potential to simulate aspects of human cognition and behavior. The paper delves into the capabilities of these models to emulate human-like text generation, offering innovative tools for literature review, hypothesis generation, experimental design, experimental subjects, data analysis, academic writing, and peer review in psychology. While LLMs are essential in advancing research methodologies in psychology, the paper also cautions about their technical and ethical challenges. There are issues like data privacy, the ethical implications of using LLMs in psychological research, and the need for a deeper understanding of these models' limitations. Researchers should responsibly use LLMs in psychological studies, adhering to ethical standards and considering the potential consequences of deploying these technologies in sensitive areas. Overall, the article provides a comprehensive overview of the current state of LLMs in psychology, exploring potential benefits and challenges. It serves as a call to action for researchers to leverage LLLs' advantages responsibly while addressing associated risks.
Natural Language Processing (NLP) is now a cornerstone of requirements automation. One compelling factor behind the growing adoption of NLP in Requirements Engineering (RE) is the prevalent use of natural language (NL) for specifying requirements in industry. NLP techniques are commonly used for automatically classifying requirements, extracting important information, e.g., domain models and glossary terms, and performing quality assurance tasks, such as ambiguity handling and completeness checking. With so many different NLP solution strategies available and the possibility of applying machine learning alongside, it can be challenging to choose the right strategy for a specific RE task and to evaluate the resulting solution in an empirically rigorous manner. This book chapter presents guidelines for the selection of NLP techniques as well as for their evaluation in the context of RE. In particular, we discuss how to choose among different strategies such as traditional NLP, feature-based machine learning, and language-model-based methods. Our ultimate hope for this chapter is to serve as a stepping stone, assisting newcomers to NLP4RE in quickly initiating themselves into the NLP technologies most pertinent to the RE field.
In this paper, we validate the performance of the a sensor fusion-based Global Navigation Satellite System (GNSS) spoofing attack detection framework for Autonomous Vehicles (AVs). To collect data, a vehicle equipped with a GNSS receiver, along with Inertial Measurement Unit (IMU) is used. The detection framework incorporates two strategies: The first strategy involves comparing the predicted location shift, which is the distance traveled between two consecutive timestamps, with the inertial sensor-based location shift. For this purpose, data from low-cost in-vehicle inertial sensors such as the accelerometer and gyroscope sensor are fused and fed into a long short-term memory (LSTM) neural network. The second strategy employs a Random-Forest supervised machine learning model to detect and classify turns, distinguishing between left and right turns using the output from the steering angle sensor. In experiments, two types of spoofing attack models: turn-by-turn and wrong turn are simulated. These spoofing attacks are modeled as SQL injection attacks, where, upon successful implementation, the navigation system perceives injected spoofed location information as legitimate while being unable to detect legitimate GNSS signals. Importantly, the IMU data remains uncompromised throughout the spoofing attack. To test the effectiveness of the detection framework, experiments are conducted in Tuscaloosa, AL, mimicking urban road structures. The results demonstrate the framework's ability to detect various sophisticated GNSS spoofing attacks, even including slow position drifting attacks. Overall, the experimental results showcase the robustness and efficacy of the sensor fusion-based spoofing attack detection approach in safeguarding AVs against GNSS spoofing threats.
This paper considers a multi-group multicasting scenario facilitated by a reconfigurable intelligent surface (RIS). We propose a fast and scalable algorithm for the joint design of the base station (BS) multicast beamforming and the RIS passive beamforming to minimize the transmit power subject to the quality-of-service (QoS) constraints. By exploring the structure of the joint optimization problem, we show that this QoS problem can be broken into a BS multicast QoS subproblem and an RIS max-min-fair (MMF) multicast subproblem, which are solved alternatingly. In our proposed algorithm, we utilize the optimal multicast beamforming structure to obtain the BS beamformers efficiently. Furthermore, we reformulate the challenging RIS multicast subproblem and employ a first-order projected subgradient algorithm (PSA) to solve it, which yields closed-form updates. Simulation results show the efficacy of our proposed algorithm in performance and computational cost compared to other alternative methods.
Large Language Models (LLMs), particularly those similar to ChatGPT, have significantly influenced the field of Natural Language Processing (NLP). While these models excel in general language tasks, their performance in domain-specific downstream tasks such as biomedical and clinical Named Entity Recognition (NER), Relation Extraction (RE), and Medical Natural Language Inference (NLI) is still evolving. In this context, our study investigates the potential of instruction tuning for biomedical language processing, applying this technique to two general LLMs of substantial scale. We present a comprehensive, instruction-based model trained on a dataset that consists of approximately $200,000$ instruction-focused samples. This dataset represents a carefully curated compilation of existing data, meticulously adapted and reformatted to align with the specific requirements of our instruction-based tasks. This initiative represents an important step in utilising such models to achieve results on par with specialised encoder-only models like BioBERT and BioClinicalBERT for various classical biomedical NLP tasks. Our work includes an analysis of the dataset's composition and its impact on model performance, providing insights into the intricacies of instruction tuning. By sharing our codes, models, and the distinctively assembled instruction-based dataset, we seek to encourage ongoing research and development in this area.
The Influence Maximization problem under the Independent Cascade model (IC) is considered. The problem asks for a minimal set of vertices to serve as "seed set" from which a maximum influence propagation is expected. New seed-set selection methods are introduced based on the notions of a $d$-packing and vertex centrality. In particular, we focus on selecting seed-vertices that are far apart and whose influence-values are the highest in their local communities. Our best results are achieved via an initial computation of a $d$-Packing followed by selecting either vertices of high degree or high centrality in their respective closed neighborhoods. This overall "Pack and Measure" approach proves highly effective as a seed selection method.
This paper gives a nearly tight characterization of the quantum communication complexity of the permutation-invariant Boolean functions. With such a characterization, we show that the quantum and randomized communication complexity of the permutation-invariant Boolean functions are quadratically equivalent (up to a logarithmic factor). Our results extend a recent line of research regarding query complexity \cite{AA14, Cha19, BCG+20} to communication complexity, showing symmetry prevents exponential quantum speedups. Furthermore, we show the Log-rank Conjecture holds for any non-trivial total permutation-invariant Boolean function. Moreover, we establish a relationship between the quantum/classical communication complexity and the approximate rank of permutation-invariant Boolean functions. This implies the correctness of the Log-approximate-rank Conjecture for permutation-invariant Boolean functions in both randomized and quantum settings (up to a logarithmic factor).
The characteristics and interpretability of data become more abstract and complex as the dimensionality increases. Common patterns and relationships that hold in in low-dimensional space may fail to hold in higher-dimensional space. This phenomenon leads to a decreasing performance for the regression, classification or clustering models or algorithms, which is known as curse of dimensionality. Curse of dimensionality can be attributed to many causes. In this paper, we first summarize five challenges associated with manipulating high-dimensional data, and explains the potential causes for the failure of regression, classification or clustering tasks. Subsequently, we delve into two major causes of the curse of dimensionality, distance concentration and manifold effect, by performing theoretical and empirical analyses. The results demonstrate that nearest neighbor search (NNS) using three typical distance measurements, Minkowski distance, Chebyshev distance, and cosine distance, becomes meaningless as the dimensionality increases. Meanwhile, the data incorporates more redundant features, and the variance contribution of principal component analysis (PCA) is skewed towards a few dimensions. By interpreting the causes of the curse of dimensionality, we can better understand the limitations of current models and algorithms, and drive to improve the performance of data analysis and machine learning tasks in high-dimensional space.
We consider the estimation of small probabilities or other risk quantities associated with rare but catastrophic events. In the model-based literature, much of the focus has been devoted to efficient Monte Carlo computation or analytical approximation assuming the model is accurately specified. In this paper, we study a distinct direction on the propagation of model uncertainty and how it impacts the reliability of rare-event estimates. Specifically, we consider the basic setup of the exceedance of i.i.d. sum, and investigate how the lack of tail information of each input summand can affect the output probability. We argue that heavy-tailed problems are much more vulnerable to input uncertainty than light-tailed problems, reasoned through their large deviations behaviors and numerical evidence. We also investigate some approaches to quantify model errors in this problem using a combination of the bootstrap and extreme value theory, showing some positive outcomes but also uncovering some statistical challenges.
Graph Neural Networks (GNNs) have been studied from the lens of expressive power and generalization. However, their optimization properties are less well understood. We take the first step towards analyzing GNN training by studying the gradient dynamics of GNNs. First, we analyze linearized GNNs and prove that despite the non-convexity of training, convergence to a global minimum at a linear rate is guaranteed under mild assumptions that we validate on real-world graphs. Second, we study what may affect the GNNs' training speed. Our results show that the training of GNNs is implicitly accelerated by skip connections, more depth, and/or a good label distribution. Empirical results confirm that our theoretical results for linearized GNNs align with the training behavior of nonlinear GNNs. Our results provide the first theoretical support for the success of GNNs with skip connections in terms of optimization, and suggest that deep GNNs with skip connections would be promising in practice.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191