This position paper argues that achieving robustness, privacy, and efficiency simultaneously in machine learning systems is infeasible under prevailing threat models. The tension between these goals arises not from algorithmic shortcomings but from structural limitations imposed by worst-case adversarial assumptions. We advocate for a systematic research agenda aimed at formalizing the robustness-privacy-efficiency trilemma, exploring how principled relaxations of threat models can unlock better trade-offs, and designing benchmarks that expose rather than obscure the compromises made. By shifting focus from aspirational universal guarantees to context-aware system design, the machine learning community can build models that are truly appropriate for real-world deployment.
The unique capabilities of Large Language Models (LLMs), such as the natural language text generation ability, position them as strong candidates for providing explanation for recommendations. However, despite the size of the LLM, most existing models struggle to produce zero-shot explanations reliably. To address this issue, we propose a framework called Logic-Scaffolding, that combines the ideas of aspect-based explanation and chain-of-thought prompting to generate explanations through intermediate reasoning steps. In this paper, we share our experience in building the framework and present an interactive demonstration for exploring our results.
A pioneering secure transmission scheme is proposed, which harnesses movable antennas (MAs) to optimize antenna positions for augmenting the physical layer security. Particularly, an MA-enabled secure wireless system is considered, where a multi-antenna transmitter communicates with a single-antenna receiver in the presence of an eavesdropper. The beamformer and antenna positions at the transmitter are jointly optimized under two criteria: power consumption minimization and secrecy rate maximization. For each scenario, a novel suboptimal algorithm was proposed to tackle the resulting nonconvex optimization problem, capitalizing on the approaches of alternating optimization and gradient descent. Numerical results demonstrate that the proposed MA systems significantly improve physical layer security compared to various benchmark schemes relying on conventional fixed-position antennas (FPAs).
Due to the importance of satellites for society and the exponential increase in the number of objects in orbit, it is important to accurately determine the state (e.g., position and velocity) of these Resident Space Objects (RSOs) at any time and in a timely manner. State-of-the-art methodologies for initial orbit determination consist of Kalman-type filters that process sequential data over time and return the state and associated uncertainty of the object, as is the case of the Extended Kalman Filter (EKF). However, these methodologies are dependent on a good initial guess for the state vector and usually simplify the physical dynamical model, due to the difficulty of precisely modeling perturbative forces, such as atmospheric drag and solar radiation pressure. Other approaches do not require assumptions about the dynamical system, such as the trilateration method, and require simultaneous measurements, such as three measurements of range and range-rate for the particular case of trilateration. We consider the same setting of simultaneous measurements (one-shot), resorting to time delay and Doppler shift measurements. Based on recent advancements in the problem of moving target localization for sonar multistatic systems, we are able to formulate the problem of initial orbit determination as a Weighted Least Squares. With this approach, we are able to directly obtain the state of the object (position and velocity) and the associated covariance matrix from the Fisher's Information Matrix (FIM). We demonstrate that, for small noise, our estimator is able to attain the Cram\'er-Rao Lower Bound accuracy, i.e., the accuracy attained by the unbiased estimator with minimum variance. We also numerically demonstrate that our estimator is able to attain better accuracy on the state estimation than the trilateration method and returns a smaller uncertainty associated with the estimation.
The electrocardiogram (ECG) is used for diagnosis and risk stratification following myocardial infarction (MI). Women have a higher incidence of missed MI diagnosis and complications following infarction, and to address this we aim to provide quantitative information on sex-differences in ECG and torso-ventricular anatomy features. A novel computational automated pipeline is presented enabling the three-dimensional reconstruction of torso-ventricular anatomies for 425 post-MI subjects and 1051 healthy controls from UK Biobank clinical images. Regression models were created relating torso-ventricular and ECG parameters. For post-MI women, the heart is positioned more posteriorly and vertically, than in men (with healthy women yet more vertical). Post-MI women exhibit less QRS prolongation, requiring 27% more prolongation than men to exceed 120ms. Only half of the sex difference in QRS is associated with smaller female cavities. Lower STj amplitude in women is striking, associated with smaller ventricles, but also more superior and posterior cardiac position. Post-MI, T wave amplitude and R axis deviations are strongly associated with a more posterior and horizontal cardiac position in women (but not in men). Our study highlights the need to quantify sex differences in anatomical features, their implications in ECG interpretation, and the application of clinical ECG thresholds in post-MI.
Reconfigurable intelligent surface (RIS)-assisted communication systems have been extensively studied for providing high-precision location services. However, most studies have overlooked the impact of carrier frequency offset (CFO) and phase noise (PN) resulting from hardware impairments on localization. This paper presents a novel, alternating optimization (AO)-based algorithm to jointly estimate the CFO, PN, and user equipment (UE) position in orthogonal frequency division multiplexing (OFDM) systems, where, provided the UE position, closed-form expressions for the CFO and PN are derived per iteration, significantly reducing the complexity and enhancing the stability of the algorithm. Another important aspect is a new RIS phase shift optimization algorithm developed to minimize the analytical lower bound of localization accuracy, hence benefiting localization. The semidefinite relaxation method and Schur complement are utilized to convexify this challenging non-convex optimization problem to a semidefinite program. Simulations demonstrate the effectiveness of the proposed algorithms, with the localization accuracy enhanced by two orders of magnitude. The localization accuracy of the proposed algorithm is close to the analytical lower bound, with a root mean square error of lower than $\rm 10^{-2} \: m$.
This paper investigates the localization problem of a network in 2-D and 3-D spaces given the positions of anchor nodes in a global frame and inter-node relative measurements in local coordinate frames. It is assumed that the local frames of different nodes have different unknown orientations. First, an angle-displacement rigidity theory is developed, which can be used to localize all the free nodes by the known positions of the anchor nodes and local relative measurements (local relative position, distance, local relative bearing, angle, or ratio-of-distance measurements). Then, necessary and sufficient conditions for network localizability are given. Finally, a distributed network localization protocol is proposed, which can globally estimate the locations of all the free nodes of a network if the network is infinitesimally angle-displacement rigid. The proposed method unifies local-relative-position-based, distance-based, local-relative-bearing-based, angle-based, and ratio-of-distance-based distributed network localization approaches. The novelty of this work is that the proposed method can be applied in both generic and non-generic configurations with an unknown global coordinate frame in both 2-D and 3-D spaces.
This paper studies 3-D distributed network localization using mixed types of local relative measurements. Each node holds a local coordinate frame without a common orientation and can only measure one type of information (relative position, distance, relative bearing, angle, or ratio-of-distance measurements) about its neighboring nodes in its local coordinate frame. A novel rigidity-theory-based distributed localization is developed to overcome the challenge due to the absence of a global coordinate frame. The main idea is to construct displacement constraints for the positions of the nodes by using mixed local relative measurements. Then, a linear distributed localization algorithm is proposed for each free node to estimate its position by solving the displacement constraints. The algebraic condition and graph condition are obtained to guarantee the global convergence of the proposed distributed localization algorithm.
This paper investigates user equipment (UE) assisted device-free networked sensing in the sixth-generation (6G) integrated sensing and communication (ISAC) system, where one base station (BS) and multiple UEs, such as unmanned aerial vehicles (UAVs), serve as anchors to cooperatively localize multiple passive targets based on the range information. Three challenges arise from the above scheme. First, the UEs are not perfectly synchronized with the BSs. Second, the UE (anchor) positions are usually estimated by the Global Positioning System (GPS) and subject to unknown errors. Third, data association is challenging, since it is hard for each anchor to associate each rang estimation to the right target under device-free sensing. We first tackle the above three challenges under a passive UE based sensing mode, where UEs only passively hear the signals over the BS-target-UE paths. A two-phase UE assisted localization protocol is proposed. In Phase I, we design an efficient method to accurately estimate the ranges from the BS to the targets and those from the BS to the targets to the UEs in the presence of synchronization errors between the BS and the UEs. In Phase II, an efficient algorithm is proposed to localize the targets via jointly removing the UEs with quite inaccurate position information from the anchor set and matching the estimated ranges at the BS and the remaining UEs with the targets. Next, we also consider an active UE based sensing mode, where the UEs can actively emit signals to obtain additional range information from them to the targets. We show that this additional range information can be utilized to significantly reduce the complexity of Phase II in the aforementioned two-phase localization protocol. Numerical results show that our proposed UE assisted networked sensing scheme can achieve very high localization accuracy.
We argue that interpretations of machine learning (ML) models or the model-building process can be seen as a form of sensitivity analysis (SA), a general methodology used to explain complex systems in many fields such as environmental modeling, engineering, or economics. We address both researchers and practitioners, calling attention to the benefits of a unified SA-based view of explanations in ML and the necessity to fully credit related work. We bridge the gap between both fields by formally describing how (a) the ML process is a system suitable for SA, (b) how existing ML interpretation methods relate to this perspective, and (c) how other SA techniques could be applied to ML.
The maturation of cognition, from introspection to understanding others, has long been a hallmark of human development. This position paper posits that for AI systems to truly emulate or approach human-like interactions, especially within multifaceted environments populated with diverse agents, they must first achieve an in-depth and nuanced understanding of self. Drawing parallels with the human developmental trajectory from self-awareness to mentalizing (also called theory of mind), the paper argues that the quality of an autonomous agent's introspective capabilities of self are crucial in mirroring quality human-like understandings of other agents. While counterarguments emphasize practicality, computational efficiency, and ethical concerns, this position proposes a development approach, blending algorithmic considerations of self-referential processing. Ultimately, the vision set forth is not merely of machines that compute but of entities that introspect, empathize, and understand, harmonizing with the complex compositions of human cognition.
We present a solution of the swing-up and balance task for the pendubot and acrobot for the participation in the AI Olympics competition at IJCAI 2023. Our solution is based on the Soft Actor Crtic (SAC) reinforcement learning (RL) algorithm for training a policy for the swing-up and entering the region of attraction of a linear quadratic regulator(LQR) controller for stabilizing the double pendulum at the top position. Our controller achieves competitive scores in performance and robustness for both, pendubot and acrobot, problem scenarios.
Traditional design cycles for new materials and assemblies have two fundamental drawbacks. The underlying physical relationships are often too complex to be precisely calculated and described. Aside from that, many unknown uncertainties, such as exact manufacturing parameters or materials composition, dominate the real assembly behavior. Machine learning (ML) methods overcome these fundamental limitations through data-driven learning. In addition, modern approaches can specifically increase system knowledge. Representation Learning allows the physical, and if necessary, even symbolic interpretation of the learned solution. In this way, the most complex physical relationships can be considered and quickly described. Furthermore, generative ML approaches can synthesize possible morphologies of the materials based on defined conditions to visualize the effects of uncertainties. This modern approach accelerates the design process for new materials and enables the prediction and interpretation of realistic materials behavior.
The role of wireless communications in various domains of intelligent transportation systems is significant; it is evident that dependable message exchange between nodes (cars, bikes, pedestrians, infrastructure, etc.) has to be guaranteed to fulfill the stringent requirements for future transportation systems. A precise site-specific digital twin is seen as a key enabler for the cost-effective development and validation of future vehicular communication systems. Furthermore, achieving a realistic digital twin for dependable wireless communications requires accurate measurement, modeling, and emulation of wireless communication channels. However, contemporary approaches in these domains are not efficient enough to satisfy the foreseen needs. In this position paper, we overview the current solutions, indicate their limitations, and discuss the most prospective paths for future investigation.
In the context of single-base station (BS) non-line-of-sight (NLoS) single-epoch localization with the aid of a reflective reconfigurable intelligent surface (RIS), this paper introduces a novel three-step algorithm that jointly estimates the position and velocity of a mobile user equipment (UE), while compensating for the Doppler effects observed in near-field (NF) at the RIS elements over the short transmission duration of a sequence of downlink (DL) pilot symbols. First, a low-complexity initialization procedure is proposed, relying in part on far-field (FF) approximation and a static user assumption. Then, an alternating optimization procedure is designed to iteratively refine the velocity and position estimates, as well as the channel gain. The refinement routines leverage small angle approximations and the linearization of the RIS response, accounting for both NF and mobility effects. We evaluate the performance of the proposed algorithm through extensive simulations under diverse operating conditions with regard to signal-to-noise ratio (SNR), UE mobility, uncontrolled multipath and RIS-UE distance. Our results reveal remarkable performance improvements over the state-of-the-art (SoTA) mobility-agnostic benchmark algorithm, while indicating convergence of the proposed algorithm to respective theoretical bounds on position and velocity estimation.
Neural radiance field (NeRF) has achieved great success in novel view synthesis and 3D representation for static scenarios. Existing dynamic NeRFs usually exploit a locally dense grid to fit the deformation field; however, they fail to capture the global dynamics and concomitantly yield models of heavy parameters. We observe that the 4D space is inherently sparse. Firstly, the deformation field is sparse in spatial but dense in temporal due to the continuity of of motion. Secondly, the radiance field is only valid on the surface of the underlying scene, usually occupying a small fraction of the whole space. We thus propose to represent the 4D scene using a learnable sparse latent space, a.k.a. SLS4D. Specifically, SLS4D first uses dense learnable time slot features to depict the temporal space, from which the deformation field is fitted with linear multi-layer perceptions (MLP) to predict the displacement of a 3D position at any time. It then learns the spatial features of a 3D position using another sparse latent space. This is achieved by learning the adaptive weights of each latent code with the attention mechanism. Extensive experiments demonstrate the effectiveness of our SLS4D: it achieves the best 4D novel view synthesis using only about $6\%$ parameters of the most recent work.
Great progress has been made in learning-based object detection methods in the last decade. Two-stage detectors often have higher detection accuracy than one-stage detectors, due to the use of region of interest (RoI) feature extractors which extract transformation-invariant RoI features for different RoI proposals, making refinement of bounding boxes and prediction of object categories more robust and accurate. However, previous RoI feature extractors can only extract invariant features under limited transformations. In this paper, we propose a novel RoI feature extractor, termed Semantic RoI Align (SRA), which is capable of extracting invariant RoI features under a variety of transformations for two-stage detectors. Specifically, we propose a semantic attention module to adaptively determine different sampling areas by leveraging the global and local semantic relationship within the RoI. We also propose a Dynamic Feature Sampler which dynamically samples features based on the RoI aspect ratio to enhance the efficiency of SRA, and a new position embedding, \ie Area Embedding, to provide more accurate position information for SRA through an improved sampling area representation. Experiments show that our model significantly outperforms baseline models with slight computational overhead. In addition, it shows excellent generalization ability and can be used to improve performance with various state-of-the-art backbones and detection methods.
Motivated by the challenge of achieving rapid learning in physical environments, this paper presents the development and training of a robotic system designed to navigate and solve a labyrinth game using model-based reinforcement learning techniques. The method involves extracting low-dimensional observations from camera images, along with a cropped and rectified image patch centered on the current position within the labyrinth, providing valuable information about the labyrinth layout. The learning of a control policy is performed purely on the physical system using model-based reinforcement learning, where the progress along the labyrinth's path serves as a reward signal. Additionally, we exploit the system's inherent symmetries to augment the training data. Consequently, our approach learns to successfully solve a popular real-world labyrinth game in record time, with only 5 hours of real-world training data.
Zero-shot voice conversion (VC) aims to transfer the source speaker timbre to arbitrary unseen target speaker timbre, while keeping the linguistic content unchanged. Although the voice of generated speech can be controlled by providing the speaker embedding of the target speaker, the speaker similarity still lags behind the ground truth recordings. In this paper, we propose SEF-VC, a speaker embedding free voice conversion model, which is designed to learn and incorporate speaker timbre from reference speech via a powerful position-agnostic cross-attention mechanism, and then reconstruct waveform from HuBERT semantic tokens in a non-autoregressive manner. The concise design of SEF-VC enhances its training stability and voice conversion performance. Objective and subjective evaluations demonstrate the superiority of SEF-VC to generate high-quality speech with better similarity to target reference than strong zero-shot VC baselines, even for very short reference speeches.
This paper addresses an uplink localization problem in which a base station (BS) aims to locate a remote user with the help of reconfigurable intelligent surfaces (RISs). We propose a strategy in which the user transmits pilots sequentially and the BS adaptively adjusts the sensing vectors, including the BS beamforming vector and multiple RIS reflection coefficients based on the observations already made, to eventually produce an estimated user position. This is a challenging active sensing problem for which finding an optimal solution involves searching through a complicated functional space whose dimension increases with the number of measurements. We show that the long short-term memory (LSTM) network can be used to exploit the latent temporal correlation between measurements to automatically construct scalable state vectors. Subsequently, the state vector is mapped to the sensing vectors for the next time frame via a deep neural network (DNN). A final DNN is used to map the state vector to the estimated user position. Numerical result illustrates the advantage of the active sensing design as compared to non-active sensing methods. The proposed solution produces interpretable results and is generalizable in the number of sensing stages. Remarkably, we show that a network with one BS and multiple RISs can outperform a comparable setting with multiple BSs.