Urban railway systems increasingly rely on communication based train control (CBTC) systems, where optimal deployment of access points (APs) in tunnels is critical for robust wireless coverage. Traditional methods, such as empirical model-based optimization algorithms, are hindered by excessive measurement requirements and suboptimal solutions, while machine learning (ML) approaches often struggle with complex tunnel environments. This paper proposes a deep reinforcement learning (DRL) driven framework that integrates parabolic wave equation (PWE) channel modeling, conditional generative adversarial network (cGAN) based data augmentation, and a dueling deep Q network (Dueling DQN) for AP placement optimization. The PWE method generates high-fidelity path loss distributions for a subset of AP positions, which are then expanded by the cGAN to create high resolution path loss maps for all candidate positions, significantly reducing simulation costs while maintaining physical accuracy. In the DRL framework, the state space captures AP positions and coverage, the action space defines AP adjustments, and the reward function encourages signal improvement while penalizing deployment costs. The dueling DQN enhances convergence speed and exploration exploitation balance, increasing the likelihood of reaching optimal configurations. Comparative experiments show that the proposed method outperforms a conventional Hooke Jeeves optimizer and traditional DQN, delivering AP configurations with higher average received power, better worst-case coverage, and improved computational efficiency. This work integrates high-fidelity electromagnetic simulation, generative modeling, and AI-driven optimization, offering a scalable and data-efficient solution for next-generation CBTC systems in complex tunnel environments.

Geofences have attracted significant attention in the design of spatial and virtual regions for managing and engaging spatiotemporal events. By using geofences to monitor human activity across their boundaries, content providers can create spatially triggered events that include notifications about points of interest within a geofence by pushing spatial information to the devices of users. Traditionally, geofences were hand-crafted by providers. In addition to the hand-crafted approach, recent advances in collecting human mobility data through mobile devices can accelerate the automatic and data-driven design of geofences, also known as the geofence design problem. Previous approaches assume circular shapes; thus, their flexibility is insufficient, and they can only handle geofence-based applications for large areas with coarse resolutions. A challenge with using circular geofences in urban and high-resolution areas is that they often overlap and fail to align with political district boundaries and road segments, such as one-way streets and median barriers. In this study, we address the problem of extracting arbitrary shapes as geofences from human mobility data to mitigate this problem. In our formulation, we cast the existing optimization problems for circular geofences to 0-1 integer programming problems to represent arbitrary shapes. Although 0-1 integer programming problems are computationally hard, formulating them as quadratic (unconstrained) binary optimization problems enables efficient approximation of optimal solutions, because this allows the use of specialized quadratic solvers, such as the quantum annealing, and other state-of-the-art algorithms. We then develop and compare different formulation methods to extract discrete geofences. We confirmed that our new modeling approach enables flexible geofence design.

Video Joint Embedding Predictive Architectures (V-JEPA) learn generalizable off-the-shelf video representation by predicting masked regions in latent space with an exponential moving average (EMA)-updated teacher. While EMA prevents representation collapse, it complicates scalable model selection and couples teacher and student architectures. We revisit masked-latent prediction and show that a frozen teacher suffices. Concretely, we (i) train a target encoder with a simple pixel-reconstruction objective under V-JEPA masking, then (ii) freeze it and train a student to predict the teacher's latents on masked regions. This leads to a two-stage, unregularized scheme that we refer to as SALT (Static-teacher Asymmetric Latent Training). SALT decouples optimization into pixel reconstruction (teacher) and masked latent prediction (student), increasing transparency, efficiency, and scalability while preserving the ability of representation to generalize under frozen evaluation. Empirically, our student models outperform recently proposed V-JEPA 2 encoders under frozen backbone evaluation across diverse benchmarks. They are also more compute-optimal: at matched pretraining FLOPs, our method achieves higher probing accuracy, and its scaling curves dominate V-JEPA's accuracy-FLOPs Pareto frontier. Finally, we find that student quality is remarkably robust to teacher quality: high-performing students emerge even with small, sub-optimal teachers. This points to a compute budget allocation that should overwhelmingly favor the student. These results position SALT as a simple, scalable, and compute-efficient alternative to EMA-based self-distillation for video representation learning.

Prompt optimization and fine-tuning are two major approaches to improve the performance of Large Language Models (LLMs). They enhance the capabilities of LLMs from complementary perspectives: the former through explicit natural language, and the latter through implicit parameter updates. However, prior work has typically studied them in isolation, leaving their synergistic potential largely underexplored. To bridge this gap, in this paper, we introduce MetaTuner, a novel framework that jointly integrates prompt optimization and fine-tuning for LLM training. Specifically, we introduce two neural networks to generate prompts and parameters, respectively, while allowing them to share a common bottom encoding layer to enable knowledge sharing. By the guidance of the final supervised signals, our framework is optimized to discover the optimal combinations between the prompts and parameters. Given that prompt learning involves discrete optimization while fine-tuning operates in a continuous parameter space, we design a supervised regularization loss to train our framework effectively. Extensive experiments across diverse benchmarks show that our method consistently outperforms the baselines.

Abstract-- Real-time dense scene reconstruction during unstable camera motions is crucial for robotics, yet current RGB-D SLAM systems fail when cameras experience large viewpoint changes, fast motions, or sudden shaking. Classical optimization-based methods deliver high accuracy but fail with poor initialization during large motions, while learning-based approaches provide robustness but lack sufficient accuracy for dense reconstruction. We address this challenge through a combination of learning-based initialization with optimization-based refinement. Our method employs a camera pose regression network to predict metric-aware relative poses from consecutive RGB-D frames, which serve as reliable starting points for a randomized optimization algorithm that further aligns depth images with the scene geometry. Extensive experiments demonstrate promising results: our approach outperforms the best competitor on challenging benchmarks, while maintaining comparable accuracy on stable motion sequences. The system operates in real-time, showcasing that combining simple and principled techniques can achieve both robustness for unstable motions and accuracy for dense reconstruction. I. INTRODUCTION Real-time camera tracking and dense scene reconstruction are fundamental problems in robotics and computer vision. For autonomous robots, handling unstable camera motions is both challenging and critical. Current RGB-D SLAM (Simultaneous Localization and Mapping) systems perform well in controlled environments with smooth, typically slow camera movements.

Active vision (AV) has been in the spotlight of robotics research due to its emergence in numerous applications including agricultural tasks such as precision crop monitoring and autonomous harvesting to list a few. A major AV problem that gained popularity is the 3D reconstruction of targeted environments using 2D images from diverse viewpoints. While collecting and processing a large number of arbitrarily captured 2D images can be arduous in many practical scenarios, a more efficient solution involves optimizing the placement of available cameras in 3D space to capture fewer, yet more informative, images that provide sufficient visual information for effective reconstruction of the environment of interest. This process termed as view planning (VP), can be markedly challenged (i) by noise emerging in the location of the cameras and/or in the extracted images, and (ii) by the need to generalize well in other unknown similar agricultural environments without need for re-optimizing or re-training. To cope with these challenges, the present work presents a novel VP framework that considers a reconstruction quality-based optimization formulation that relies on the notion of `structure-from-motion' to reconstruct the 3D structure of the sought environment from the selected 2D images. With no analytic expression of the optimization function and with costly function evaluations, a Bayesian optimization approach is proposed to efficiently carry out the VP process using only a few function evaluations, while accounting for different noise cases. Numerical tests on both simulated and real agricultural settings signify the benefits of the advocated VP approach in efficiently estimating the optimal camera placement to accurately reconstruct 3D environments of interest, and generalize well on similar unknown environments.

Stochastic first-order optimization methods, such as Stochastic Gradient Descent ( SGD) (Robbins and Monro, 1951), AdaGrad (Streeter and McMahan, 2010; Duchi et al., 2011), and Adam (Kingma and Ba, 2014), are fundamental for training modern Machine Learning (ML) and Deep Learning (DL) models. However, these methods are often enhanced with additional algorithmic techniques that play a critical role in their convergence and practical performance. Among these, gradient clipping (Pascanu et al., 2013) is one of the most widely used and well-studied approaches. In recent years, substantial efforts have been made to theoretically understand the advantages of gradient clipping and its impact on the convergence of stochastic optimization algorithms. In particular, gradient clipping is a key component in managing heavy-tailed noise, which commonly arises in the training of language models on textual data (Zhang et al., 2020b), in the training of GANs (Goodfellow et al., 2014; Gorbunov et al., 2022), and even in simpler tasks such as image classification (S im sekli et al., 2019). This approach is primarily analyzed through the lens of high-probability convergence, as such guarantees provide a more accurate reflection of the actual behavior of optimization methods compared to their more conventional in-expectation counterparts (Gorbunov et al., 2020).

We investigate a general approach for improving user prompts in text-to-image (T2I) diffusion models by finding prompts that maximize a reward function specified at test-time. Although diverse reward models are used for evaluating image generation, existing automated prompt engineering methods typically target specific reward configurations. Consequently, these specialized designs exhibit suboptimal performance when applied to new prompt engineering scenarios involving different reward models. To address this limitation, we introduce RATTPO (Reward-Agnostic Test-Time Prompt Optimization), a flexible test-time optimization method applicable across various reward scenarios without modification. RATTPO iteratively searches for optimized prompts by querying large language models (LLMs) \textit{without} requiring reward-specific task descriptions. Instead, it uses the optimization trajectory and a novel reward-aware feedback signal (termed a "hint") as context. Empirical results demonstrate the versatility of RATTPO, effectively enhancing user prompts across diverse reward setups that assess various generation aspects, such as aesthetics, general human preference, or spatial relationships between objects. RATTPO surpasses other test-time search baselines in search efficiency, running 4.8 times faster than naive reward-agnostic test-time search baseline on average. Furthermore, with sufficient inference budget, it can achieve comparable performance to learning-based baselines that require reward-specific fine-tuning. The code is available at https://github.com/seminkim/RATTPO.

Feature selection removes redundant features to enhanc performance and computational efficiency in downstream tasks. Existing works often struggle to capture complex feature interactions and adapt to diverse scenarios. Recent advances in this domain have incorporated generative intelligence to address these drawbacks by uncovering intricate relationships between features. However, two key limitations remain: 1) embedding feature subsets in a continuous space is challenging due to permutation sensitivity, as changes in feature order can introduce biases and weaken the embedding learning process; 2) gradient-based search in the embedding space assumes convexity, which is rarely guaranteed, leading to reduced search effectiveness and suboptimal subsets. To address these limitations, we propose a new framework that can: 1) preserve feature subset knowledge in a continuous embedding space while ensuring permutation invariance; 2) effectively explore the embedding space without relying on strong convex assumptions. For the first objective, we develop an encoder-decoder paradigm to preserve feature selection knowledge into a continuous embedding space. This paradigm captures feature interactions through pairwise relationships within the subset, removing the influence of feature order on the embedding. Moreover, an inducing point mechanism is introduced to accelerate pairwise relationship computations. For the second objective, we employ a policy-based reinforcement learning (RL) approach to guide the exploration of the embedding space. The RL agent effectively navigates the space by balancing multiple objectives. By prioritizing high-potential regions adaptively and eliminating the reliance on convexity assumptions, the RL agent effectively reduces the risk of converging to local optima. Extensive experiments demonstrate the effectiveness, efficiency, robustness and explicitness of our model.

This study explores the performance of the random Gaussian smoothing Zeroth-Order ExtraGradient (ZO-EG) scheme considering \Af{deterministic} min-max optimisation problems with possibly NonConvex-NonConcave (NC-NC) objective functions. We consider both unconstrained and constrained, differentiable and non-differentiable settings. We discuss the min-max problem from the point of view of variational inequalities. For the unconstrained problem, we establish the convergence of the ZO-EG algorithm to the neighbourhood of an $ε$-stationary point of the NC-NC objective function, whose radius can be controlled under a variance reduction scheme, along with its complexity. For the constrained problem, we introduce the new notion of proximal variational inequalities and give examples of functions satisfying this property. Moreover, we prove analogous results to the unconstrained case for the constrained problem. For the non-differentiable case, we prove the convergence of the ZO-EG algorithm to a neighbourhood of an $ε$-stationary point of the smoothed version of the objective function, where the radius of the neighbourhood can be controlled, which can be related to the ($δ,ε$)-Goldstein stationary point of the original objective function.