In this paper, we present electromagnetic signal and information theory (ESIT). ESIT is an interdisciplinary scientific discipline, which amalgamates electromagnetic theory, signal processing theory, and information theory. ESIT is aimed at studying and designing physically consistent communication schemes for the transmission and processing of information in communication networks. In simple terms, ESIT can be defined as physics-aware information theory and signal processing for communications. We consider three relevant problems in contemporary communication theory, and we show how they can be tackled under the lenses of ESIT. Specifically, we focus on (i) the theoretical and practical motivations behind antenna designs based on subwavelength radiating elements and interdistances; (ii) the modeling and role played by the electromagnetic mutual coupling, and the appropriateness of multiport network theory for modeling it; and (iii) the analytical tools for unveiling the performance limits and realizing spatial multiplexing in near field, line-of-sight, channels. To exemplify the role played by ESIT and the need for electromagnetic consistency, we consider case studies related to reconfigurable intelligent surfaces and holographic surfaces, and we highlight the inconsistencies of widely utilized communication models, as opposed to communication models that originate from first electromagnetic principles.
相關內容
Numerical reasoning is an essential ability for NLP systems to handle numeric information. Recent research indicates that fine-tuning a small-scale model to learn generating reasoning processes alongside answers can significantly enhance performance. However, current methods have the limitation that most methods generate reasoning processes with large language models (LLMs), which are "unreliable" since such processes could contain information unrelated to the answer. To address this limitation, we introduce Enhancing NumeriCal reasOning with Reliable procEsses (Encore), which derives the reliable reasoning process by decomposing the answer formula, ensuring which fully supports the answer. Nevertheless, models could lack enough data to learn the reasoning process generation adequately, since our method generates only one single reasoning process for one formula. To overcome this difficulty, we present a series of pre-training tasks to help models learn the reasoning process generation with synthesized data. The experiments show that Encore yields improvement on all five experimental datasets with an average of 1.8%, proving the effectiveness of our method.
Advances towards more faithful and traceable answers of Large Language Models (LLMs) are crucial for various research and practical endeavors. One avenue in reaching this goal is basing the answers on reliable sources. However, this Evidence-Based QA has proven to work insufficiently with LLMs in terms of citing the correct sources (source quality) and truthfully representing the information within sources (answer attributability). In this work, we systematically investigate how to robustly fine-tune LLMs for better source quality and answer attributability. Specifically, we introduce a data generation pipeline with automated data quality filters, which can synthesize diversified high-quality training and testing data at scale. We further introduce four test sets to benchmark the robustness of fine-tuned specialist models. Extensive evaluation shows that fine-tuning on synthetic data improves performance on both in- and out-of-distribution. Furthermore, we show that data quality, which can be drastically improved by proposed quality filters, matters more than quantity in improving Evidence-Based QA.
Reconfigurable intelligent surface (RIS) has great potential to improve the performance of integrated sensing and communication (ISAC) systems, especially in scenarios where line-of-sight paths between the base station and users are blocked. However, the spectral efficiency (SE) of RIS-aided ISAC uplink transmissions may be drastically reduced by the heavy burden of pilot overhead for realizing sensing capabilities. In this paper, we tackle this bottleneck by proposing a superimposed symbol scheme, which superimposes sensing pilots onto data symbols over the same time-frequency resources. Specifically, we develop a structure-aware sparse Bayesian learning framework, where decoded data symbols serve as side information to enhance sensing performance and increase SE. To meet the low-latency requirements of emerging ISAC applications, we further propose a low-complexity simultaneous communication and localization algorithm for multiple users. This algorithm employs the unitary approximate message passing in the Bayesian learning framework for initial angle estimate, followed by iterative refinements through reduced-dimension matrix calculations. Moreover, the sparse code multiple access technology is incorporated into this iterative framework for accurate data detection which also facilitates localization. Numerical results show that the proposed superimposed symbol-based scheme empowered by the developed algorithm can achieve centimeter-level localization while attaining up to $96\%$ of the SE of conventional communications without sensing capabilities. Moreover, compared to other typical ISAC schemes, the proposed superimposed symbol scheme can provide an effective throughput improvement over $133\%$.
In this paper, we propose an efficient multi-stage algorithm for non-adaptive Group Testing (GT) with general correlated prior statistics. The proposed solution can be applied to any correlated statistical prior represented in trellis, e.g., finite state machines and Markov processes. We introduce a variation of List Viterbi Algorithm (LVA) to enable accurate recovery using much fewer tests than objectives, which efficiently gains from the correlated prior statistics structure. Our numerical results demonstrate that the proposed Multi-Stage GT (MSGT) algorithm can obtain the optimal Maximum A Posteriori (MAP) performance with feasible complexity in practical regimes, such as with COVID-19 and sparse signal recovery applications, and reduce in the scenarios tested the number of pooled tests by at least $25\%$ compared to existing classical low complexity GT algorithms. Moreover, we analytically characterize the complexity of the proposed MSGT algorithm that guarantees its efficiency.
Background: Pose estimation of rigid objects is a practical challenge in optical metrology and computer vision. This paper presents a novel stochastic-geometrical modeling framework for object pose estimation based on observing multiple feature points. Methods: This framework utilizes mixture models for feature point densities in object space and for interpreting real measurements. Advantages are the avoidance to resolve individual feature correspondences and to incorporate correct stochastic dependencies in multi-view applications. First, the general modeling framework is presented, second, a general algorithm for pose estimation is derived, and third, two example models (camera and lateration setup) are presented. Results: Numerical experiments show the effectiveness of this modeling and general algorithm by presenting four simulation scenarios for three observation systems, including the dependence on measurement resolution, object deformations and measurement noise. Probabilistic modeling utilizing mixture models shows the potential for accurate and robust pose estimations while avoiding the correspondence problem.
Imitation Learning (IL) is a promising paradigm for teaching robots to perform novel tasks using demonstrations. Most existing approaches for IL utilize neural networks (NN), however, these methods suffer from several well-known limitations: they 1) require large amounts of training data, 2) are hard to interpret, and 3) are hard to repair and adapt. There is an emerging interest in programmatic imitation learning (PIL), which offers significant promise in addressing the above limitations. In PIL, the learned policy is represented in a programming language, making it amenable to interpretation and repair. However, state-of-the-art PIL algorithms assume access to action labels and struggle to learn from noisy real-world demonstrations. In this paper, we propose PLUNDER, a novel PIL algorithm that integrates a probabilistic program synthesizer in an iterative Expectation-Maximization (EM) framework to address these shortcomings. Unlike existing PIL approaches, PLUNDER synthesizes probabilistic programmatic policies that are particularly well-suited for modeling the uncertainties inherent in real-world demonstrations. Our approach leverages an EM loop to simultaneously infer the missing action labels and the most likely probabilistic policy. We benchmark PLUNDER against several established IL techniques, and demonstrate its superiority across five challenging imitation learning tasks under noise. PLUNDER policies achieve 95% accuracy in matching the given demonstrations, outperforming the next best baseline by 19%. Additionally, policies generated by PLUNDER successfully complete the tasks 17% more frequently than the nearest baseline.
With the advances of data-driven machine learning research, a wide variety of prediction problems have been tackled. It has become critical to explore how machine learning and specifically deep learning methods can be exploited to analyse healthcare data. A major limitation of existing methods has been the focus on grid-like data; however, the structure of physiological recordings are often irregular and unordered which makes it difficult to conceptualise them as a matrix. As such, graph neural networks have attracted significant attention by exploiting implicit information that resides in a biological system, with interactive nodes connected by edges whose weights can be either temporal associations or anatomical junctions. In this survey, we thoroughly review the different types of graph architectures and their applications in healthcare. We provide an overview of these methods in a systematic manner, organized by their domain of application including functional connectivity, anatomical structure and electrical-based analysis. We also outline the limitations of existing techniques and discuss potential directions for future research.
In this paper, we propose a novel Feature Decomposition and Reconstruction Learning (FDRL) method for effective facial expression recognition. We view the expression information as the combination of the shared information (expression similarities) across different expressions and the unique information (expression-specific variations) for each expression. More specifically, FDRL mainly consists of two crucial networks: a Feature Decomposition Network (FDN) and a Feature Reconstruction Network (FRN). In particular, FDN first decomposes the basic features extracted from a backbone network into a set of facial action-aware latent features to model expression similarities. Then, FRN captures the intra-feature and inter-feature relationships for latent features to characterize expression-specific variations, and reconstructs the expression feature. To this end, two modules including an intra-feature relation modeling module and an inter-feature relation modeling module are developed in FRN. Experimental results on both the in-the-lab databases (including CK+, MMI, and Oulu-CASIA) and the in-the-wild databases (including RAF-DB and SFEW) show that the proposed FDRL method consistently achieves higher recognition accuracy than several state-of-the-art methods. This clearly highlights the benefit of feature decomposition and reconstruction for classifying expressions.
Deep neural networks (DNNs) are successful in many computer vision tasks. However, the most accurate DNNs require millions of parameters and operations, making them energy, computation and memory intensive. This impedes the deployment of large DNNs in low-power devices with limited compute resources. Recent research improves DNN models by reducing the memory requirement, energy consumption, and number of operations without significantly decreasing the accuracy. This paper surveys the progress of low-power deep learning and computer vision, specifically in regards to inference, and discusses the methods for compacting and accelerating DNN models. The techniques can be divided into four major categories: (1) parameter quantization and pruning, (2) compressed convolutional filters and matrix factorization, (3) network architecture search, and (4) knowledge distillation. We analyze the accuracy, advantages, disadvantages, and potential solutions to the problems with the techniques in each category. We also discuss new evaluation metrics as a guideline for future research.
Deep convolutional neural networks (CNNs) have recently achieved great success in many visual recognition tasks. However, existing deep neural network models are computationally expensive and memory intensive, hindering their deployment in devices with low memory resources or in applications with strict latency requirements. Therefore, a natural thought is to perform model compression and acceleration in deep networks without significantly decreasing the model performance. During the past few years, tremendous progress has been made in this area. In this paper, we survey the recent advanced techniques for compacting and accelerating CNNs model developed. These techniques are roughly categorized into four schemes: parameter pruning and sharing, low-rank factorization, transferred/compact convolutional filters, and knowledge distillation. Methods of parameter pruning and sharing will be described at the beginning, after that the other techniques will be introduced. For each scheme, we provide insightful analysis regarding the performance, related applications, advantages, and drawbacks etc. Then we will go through a few very recent additional successful methods, for example, dynamic capacity networks and stochastic depths networks. After that, we survey the evaluation matrix, the main datasets used for evaluating the model performance and recent benchmarking efforts. Finally, we conclude this paper, discuss remaining challenges and possible directions on this topic.
小貼士
登錄享
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191