Online ride-hailing platforms aim to deliver efficient mobility-on-demand services, often facing challenges in balancing dynamic and spatially heterogeneous supply and demand. Existing methods typically fall into two categories: reinforcement learning (RL) approaches, which suffer from data inefficiency, oversimplified modeling of real-world dynamics, and difficulty enforcing operational constraints; or decomposed online optimization methods, which rely on manually designed high-level objectives that lack awareness of low-level routing dynamics. To address this issue, we propose a novel hybrid framework that integrates large language model (LLM) with mathematical optimization in a dynamic hierarchical system: (1) it is training-free, removing the need for large-scale interaction data as in RL, and (2) it leverages LLM to bridge cognitive limitations caused by problem decomposition by adaptively generating high-level objectives. Within this framework, LLM serves as a meta-optimizer, producing semantic heuristics that guide a low-level optimizer responsible for constraint enforcement and real-time decision execution. These heuristics are refined through a closed-loop evolutionary process, driven by harmony search, which iteratively adapts the LLM prompts based on feasibility and performance feedback from the optimization layer. Extensive experiments based on scenarios derived from both the New York and Chicago taxi datasets demonstrate the effectiveness of our approach, achieving an average improvement of 16% compared to state-of-the-art baselines.

The technology of firefighting unmanned aerial vehicles (UAVs) used in fire scene reconnaissance, water disaster investigation, and rescue missions, has become a crucial asset in responding to fires and emergencies. The key to this technology is to swiftly and effectively improve the operational efficiency of UAVs. Based on the premise of rapid and effective rescue, the stability, tracking, and safety of hexacopter UAVs technology improved the design and optimization innovative design of the storage and transportation convenience structure of unmanned aerial vehicles were carried out. An innovative high-pressure ejection mechanism for fire bombs (delivery items) has been designed. So, a portable, efficient, and fast unmanned aerial vehicle structure was designed. The PX4 flight control system was used to conduct simulation experiments on the newly designed UAV, and it was found that the structural design was reasonable, and the UAV was operated smoothly and accurately during the simulation flight. This structural design and simulation analysis can provide new ideas for the design of new fire-fighting equipment.

Internet of Things (IoT) devices generate vast amounts of real-time data across diverse sectors, offering lucrative opportunities in the data trading market. This facilitates the conversion of raw data into valuable products and services, resulting in significant economic and social benefits. To issue data security problems in trading mechanisms, several solutions based on local differential privacy (LDP) provide lightweight methods for privacy preservation and efficient data exchange. However, most solutions lack effective mechanisms for longitudinal data. Moreover, LDP needs to address the data quality problem in IoT. At last, the pricing of privacy-preserving longitudinal data remains unresolved. To address these problems, we propose a privacy-preserving data trading (PPDT) scheme for IoT in this article. Specifically, to guarantee the security of longitudinal data, we utilize two perturbation techniques to accommodate data owners with varying privacy preferences. To enhance data availability, we devise a binary tree-based aggregation algorithm combined with a weighted average strategy and maximum likelihood estimation. Additionally, we derive an optimal contract that considers different levels of privacy preservation and data trading prices. In scenarios with incomplete information, the contract can provide appropriate incentives to the involved parties. Finally, we demonstrate the efficiency and effectiveness of the proposed data trading scheme through theoretical analysis and extensive experiments.

Range queries are extensively used in various Internet of Things (IoT) applications as an essential functional requirement to provide intelligent and personalized services to users. In IoT environments, diverse types of data are generated, necessitating the design of range query schemes for multidimensional data. Privacy preservation is a key concern for range queries, leading to the proposal of several privacy-preserving solutions. However, most of these solutions are either inefficient or impractical. Moreover, existing approaches often suffer from the problem of longitudinal data privacy leakage, posing a serious threat to user privacy. Although some efforts have addressed the privacy issues of longitudinal data, practical implementations have been hesitant. To tackle these challenges, we propose a local differential privacy-based (LDP) privacy-preserving scheme called the privacy-preserving longitudinal and multidimensional range query (PLMRQ) for IoT. Our scheme focuses on lightweight privacy preservation and eliminates the need for a trusted third party (TTP). First, it is designed based on a double randomizer, ensuring effective privacy preservation of longitudinal data over time. Second, to mitigate excessive noise injection, PLMRQ dynamically constructs a binary tree structure by hierarchically decomposing the entire domain. Finally, through the utilization of a post-processing technique, the mean square error is efficiently reduced. Theoretical and experimental results demonstrate that the proposed PLMRQ maintains competitive utility while rigorously satisfying$\ln {({e^{\epsilon _{1}+t\epsilon _{2}}+1}/{e^{\epsilon _{1}}+e^{t\epsilon _{2}}})}$-LDP with an upper bound of$\epsilon _{1}$and a lower bound of$\epsilon _{2}$.

Recent advances in machine learning have leveraged dramatic increases in computational power, a trend expected to continue in the future. This paper introduces the first Hyperscale Hardware Optimized Neural Architecture Search (H2O-NAS) to automatically design accurate and performant machine learning models tailored to the underlying hardware architecture. H2O-NAS consists of three key components: a new massively parallel “one-shot” search algorithm with intelligent weight sharing, which can scale to search spaces of O(10280) and handle large volumes of production traffic; hardware-optimized search spaces for diverse ML models on heterogeneous hardware; and a novel two-phase hybrid performance model and a multi-objective reward function optimized for large scale deployments.

One promising way to accelerate transformer training is to reuse small pretrained models to initialize the transformer, as their existing representation power facilitates faster model convergence. Previous works designed expansion operators to scale up pretrained models to the target model before training. Yet, model functionality is difficult to preserve when scaling a transformer in all dimensions at once. Moreover, maintaining the pretrained optimizer states for weights is critical for model scaling, whereas the new weights added during expansion lack these states in pre-trained models. To address these issues, we propose TripLe, which partially scales a model before training, while growing the rest of the new parameters during training by copying both the warmed-up weights with the optimizer states from existing weights. As such, the new parameters introduced during training will obtain their training states. Furthermore, through serializing the scaling of model width and depth, the functionality of each expansion can be preserved. We evaluate TripLe in both single-trial model scaling and multi-trial neural architecture search (NAS). Due to the fast training convergence of TripLe, the proxy accuracy from TripLe better reveals the model quality compared to from-scratch training in multi-trial NAS. Experiments show that TripLe outperforms from-scratch training and knowledge distillation (KD) in both training time and task performance. TripLe can also be combined with KD to achieve an even higher task accuracy. For NAS, the model obtained from TripLe outperforms DeiT-B in task accuracy with 69% reduction in parameter size and FLOPs.

A* algorithm and ant colony optimization (ACO) are more widely used in path planning among global path planning algorithms. The optimization process is analyzed and summarized from the principles and characteristics of the two algorithms, A* algorithm is mainly optimized in terms of point selection and improvement of heuristic function; and ACO is mainly investigated in terms of transfer probability and pheromone positive feedback for improvement and optimization. Taking a single algorithm solving complex optimization problems difficulties into consideration, a splitting strategy can be used. So that local path or intelligent path optimization algorithms are incorporated in global path planning to improve search efficiency and optimization quality.

Understanding animals' behaviors is significant for a wide range of applications. However, existing animal behavior datasets have limitations in multiple aspects, including limited numbers of animal classes, data samples and provided tasks, and also limited variations in environmental conditions and viewpoints. To address these limitations, we create a large and diverse dataset, Animal Kingdom, that provides multiple annotated tasks to enable a more thorough understanding of natural animal behaviors. The wild animal footages used in our dataset record different times of the day in extensive range of environments containing variations in backgrounds, viewpoints, illumination and weather conditions. More specifically, our dataset contains 50 hours of annotated videos to localize relevant animal behavior segments in long videos for the video grounding task, 30K video sequences for the fine-grained multi-label action recognition task, and 33K frames for the pose estimation task, which correspond to a diverse range of animals with 850 species across 6 major animal classes. Such a challenging and comprehensive dataset shall be able to facilitate the community to develop, adapt, and evaluate various types of advanced methods for animal behavior analysis. Moreover, we propose a Collaborative Action Recognition (CARe) model that learns general and specific features for action recognition with unseen new animals. This method achieves promising performance in our experiments. Our dataset can be found at https://sutdcv.github.io/Animal-Kingdom.

Human behavior understanding with unmanned aerial vehicles (UAVs) is of great significance for a wide range of applications, which simultaneously brings an urgent demand of large, challenging, and comprehensive benchmarks for the development and evaluation of UAV-based models. However, existing benchmarks have limitations in terms of the amount of captured data, types of data modalities, categories of provided tasks, and diversities of subjects and environments. Here we propose a new benchmark - UAV-Human - for human behavior understanding with UAVs, which contains 67,428 multi-modal video sequences and 119 subjects for action recognition, 22,476 frames for pose estimation, 41,290 frames and 1,144 identities for person re-identification, and 22,263 frames for attribute recognition. Our dataset was collected by a flying UAV in multiple urban and rural districts in both daytime and night-time over three months, hence covering extensive diversities w.r.t subjects, backgrounds, illuminations, weathers, occlusions, camera motions, and UAV flying attitudes. Such a comprehensive and challenging benchmark shall be able to promote the research of UAV-based human behavior understanding, including action recognition, pose estimation, re-identification, and attribute recognition. Furthermore, we propose a fisheye-based action recognition method that mitigates the distortions in fisheye videos via learning unbounded transformations guided by flat RGB videos. Experiments show the efficacy of our method on the UAV-Human dataset.

Depth estimation is a fundamental problem for light field photography applications. Numerous methods have been proposed in recent years, which either focus on crafting cost terms for more robust matching, or on analyzing the geometry of scene structures embedded in the epipolar-plane images. Significant improvements have been made in terms of overall depth estimation error; however, current state-of-the-art methods still show limitations in handling intricate occluding structures and complex scenes with multiple occlusions. To address these challenging issues, we propose a very effective depth estimation framework which focuses on regularizing the initial label confidence map and edge strength weights. Specifically, we first detect partially occluded boundary regions (POBR) via superpixel-based regularization. Series of shrinkage/reinforcement operations are then applied on the label confidence map and edge strength weights over the POBR. We show that after weight manipulations, even a low-complexity weighted least squares model can produce much better depth estimation than the state-of-the-art methods in terms of average disparity error rate, occlusion boundary precision-recall rate, and the preservation of intricate visual features.

Photography through reflective surfaces suffers from the obstruction of reflections, which deteriorates the visibility of background targets and causes challenges for subsequent computer vision applications. In this paper, we propose a novel reflection removal algorithm using light field (LF) cameras. Unlike conventional cameras, LF cameras capture extra directional information of incoming rays which enable our algorithm to remove reflections with only a single shot. We analyze the optical geometry of the background and reflection imagery in a LF camera, and generalize a set of rules that could facilitate our algorithm to differentiate the edges from different optical sources. We show that the proposed method produces significantly better reflection removal results based on the LF data as compared to traditional methods based on multiple-shot image sequences as input.