Siamese-Like Architecture for Aerial Crowd Density Estimation with Refined Counting From Real and Synthetic Data

This chapter aims to explore the application of Internet of Things (IoT) and machine learning (ML) in assessing the planning and management of smart cities and new towns. Additionally, we have discussed the possible solutions that combine IoT and ML opportunities in this field. Similarly, our results support that ML and IoT optimize resources, address environmental challenges and enhance citizen well-being by using real-time data and predictive models. However, challenges such as interoperability, cybersecurity, data protection, algorithmic bias and high costs must be addressed. Finally, we support that by adopting transparent governance and innovation, smart cities and new towns can build resilient urban ecosystems that prioritize the needs of their residents and ensure equitable development.

It was analyzed the information of databases, encryption algorithms, and information security that are the main components of a Smart City. The problem is the lack of security of public or secret data that is stored in a repository, where there is access to the citizen, local government, and central government. The objective is to define a security prototype and adopt a cryptographic algorithm to mitigate the risks of the database in the management of a Smart City. It was used the deductive method and exploratory research to analyze the information of the reference articles. It turned out a general prototype of security; an algorithm to implement a database for Smart City; an algorithm of data protection expressed in flowcharts. It was concluded that information protected with an encryption algorithm, is a support to be more efficient, reduce costs, reduce the environmental footprint and improve the management of a Smart City.

24 - Current trends and future challenges in smart waste management in smart cities

A massive wave of urbanisation has grasped both developed and developing nations in the last decade. Various studies showcase the rising trend of rapid urbanisation growth due to demographic shifts, catalysed by numerous global and local parameters. India is professed to house 50 percent of its population in urban areas by 2030. The urban citizen’s rising expectations regarding infrastructure, amenities and safety will dovetail with this phenomenon. To cater to the demands of increased urbanisation, the concept of Smart cities is being advocated as an apt solution. Even so, there is widespread ambiguity about the challenges associated with adopting smart solutions into the existing urban social form. Though there have been many tentative explanations regarding the basic framework of a smart city, there are no known validated definitions. The main notable feature is the use of information and communication technologies (ICTs) within the cyclic dynamics of a city to facilitate the smooth and cost-effective working of urban areas. This ICT infrastructure enables real-time data collection, analysis, response, and storage. This has been deemed beneficial due to the increase in efficiency of operation and management services involved with the day-to-day functioning of any urban area. Smart cities advocate using information and communication technologies (ICTs) within the dynamics of a city to enable the real-time collection, analysis, and storage of big data. This is beneficial due to the increased efficiency of operation and management services involved with an urban area’s daily functioning. One such technological intervention is the ‘drone’ or the ‘unmanned aerial vehicles (UAVs)’. UAVs have a wide variety of uses in a smart urban fabric, from geospatial integration to traffic management, surveillance, disaster response, etc. In the current nascent stage of research and development of UAVs as one of the innovative solutions for smart cities in India, questions arise regarding privacy, cost of production, technical knowledge, safety and security with their large-scale use. This paper aims to assess the applicability of UAVs in overall smart city planning and management. The feasibility analysis method is adopted to analyse the felicitousness of UAVs in a smart city’s planning and design phases. The results of the study undertaken in this paper highlight the challenges and opportunities in the planning and management of smart cities by integrating UAVs. The paper enumerates the relevance and appropriate benefits of using UAVs to plan, design, and perpetuate Indian smart cities.

The rise of electric and autonomous vehicles in smart cities poses challenges in vehicular energy management due to un-optimized consumption, inefficient grid use, and unpredictable traffic patterns. Traditional centralized machine learning models and cloud-based Energy Management Systems (EMSs) struggle with real-time adaptability, high-dimensional data processing, and data privacy risks. These issues lead to high costs, excessive energy waste, and regulatory concerns. Federated Learning (FL) offers a decentralized approach where multiple edge devices collaboratively train models without sharing raw data. This enhances privacy, reduces communication overhead, and is well-suited for smart city applications. However, FL’s black-box nature limits interpretability, reducing trust in AI-driven decisions. Explainable AI (XAI) addresses this by enhancing transparency, interpretability, and regulatory compliance. This research introduces Explainable FL (XFL) for optimizing vehicular energy management in smart cities. The proposed XFL framework integrates distributed learning with explainability techniques for interpretable and accountable decision-making. Using a real-world AEV telemetry dataset of approximately 1,219,567 records with features like speed, energy consumption, and traffic density, it employs a hierarchical FL architecture to ensure secure and decentralized learning. It efficiently analyzes real-time traffic, vehicle energy states, and grid load balancing while preserving privacy. Experimental results show that the proposed Multi-Layer Perceptron (MLP)-based global model achieves superior predictive accuracy, with R² values of 94.73% for energy consumption and 99.83% for traffic density, significantly outperforming previous methods.

This paper proposes a method to manage and utilize image and video data effectively in a smart city. Applying digital watermarking techniques, a framework for visual data security and management for smart cities is presented. In the framework, a reversible fragile or semi-fragile watermark embedded into the visual data is used to ensure trusted acquisition. Moreover, reversible metadata watermarks carrying information such as identification and other properties data is used to assist visual data management. A solution for tracing users on a large scale is presented using reversible watermarking.

An intelligent transportation system for smart city applications is regarded as a solution for reducing traffic congestion, fuel consumption, and accidents, and saving lives by levering computing, communication, control, and collected traffic data. Currently, traffic officials are not only facing a significant absence of automated and accurate traffic prediction but also making informed decisions in a real-time manner for traffic management in smart cities. This chapter proposes an automated method for counting vehicles for identifying traffic patterns and offering an adaptive traffic management system by using deep learning. The proposed approach helps to count vehicles in the intersection which helps to make an informed decision about making streets one-way or two-way adaptive traffic light systems depending not only on the number of vehicles in the intersection but also based on the time of the day. Specifically, the proposed approach detects and counts turning vehicles, and pedestrians at a traffic intersection in a real-time manner (as well as in recorded videos in case of offline processing) using deep learning. Then the collected statistics are used to make an informed decision. The proposed approach gives better results (in terms of time and accuracy) compared to state-of-the-art work.

Accurate wind‐power forecasting is essential for the successful integration of renewable energy into smart city grids. Traditional long short‐term memory (LSTM) networks face significant challenges in parameter optimization, leading to suboptimal forecasting performance. This paper introduces a novel Bernstein dynamic horned lizard optimization algorithm (BDHLOA) to optimize LSTM parameters for wind power prediction in smart cities. BDHLOA incorporates three key enhancements: a Bernstein‐assisted oppositional‐multiple learning (BOML) strategy that improves exploration–exploitation balance, Bernstein‐based adaptive differential (BAD) strategy for better solution refinement, and a dynamic drift search (DDS) mechanism that prevents premature convergence. The proposed BDHLOA‐LSTM framework is evaluated for one‐step‐ahead (10‐min horizon) wind‐power forecasting using four real‐world datasets from La Haute Borne wind turbines in France, under a rolling‐origin expanding‐window cross‐validation protocol that strictly prevents data leakage. Results demonstrate exceptional performance across all four stations, with mean R 2 = 0.9695, RMSE = 0.0011, and MAE = 0.0007. BDHLOA‐LSTM reduces MAE by 94% compared to persistence forecasting and 91% compared to autoregressive AR(24) models. Against the best competing optimizer (PSO‐LSTM), BDHLOA‐LSTM achieves 42% lower MAE on average across sites. Performance is assessed using R 2 , RMSE, MAE, symmetric mean absolute percentage error (sMAPE), mean absolute scaled error (MASE), and prediction interval coverage probability (PICP) to capture both accuracy and calibrated uncertainty. Furthermore, the proposed model is interpreted using SHapley Additive exPlanations (SHAP) technique. SHAP analysis confirms that BDHLOA‐LSTM learns physically meaningful relationships dominated by wind speed statistics and direction, rather than exploiting spurious correlations. The superior accuracy and stability of BDHLOA‐LSTM make it highly suitable for real‐time grid management and sustainable energy planning in smart cities.

Effective response strategies to earthquake disasters are crucial for disaster management in smart cities. However, in regions where earthquakes do not occur frequently, model construction may be difficult due to a lack of training data. To address this issue, there is a need for technology that can generate earthquake scenarios for response training at any location. We proposed a model for generating earthquake scenarios using an auxiliary classifier Generative Adversarial Network (AC-GAN)-based data synthesis. The proposed ACGAN model generates various earthquake scenarios by incorporating an auxiliary classifier learning process into the discriminator of GAN. Our results at borehole sensors showed that the seismic data generated by the proposed model had similar characteristics to actual data. To further validate our results, we compared the generated IM (such as PGA, PGV, and SA) with Ground Motion Prediction Equations (GMPE). Furthermore, we evaluated the potential of using the generated scenarios for earthquake early warning training. The proposed model and algorithm have significant potential in advancing seismic analysis and detection management systems, and also contribute to disaster management.

Building a smart city is not only a need of realizing sustainable urban development, but also the strategic choice of improving China’s comprehensive competitiveness. The technology of Big Data can suply strong data support for the smart city’s construction, and is of great significance to development, construction and management of smart city. This paper points out the existing problems and deficiencies in building smart city’s process, and puts forward the framework of smart city management platform on account of the technology of big data, thus maximize the value of big data, push forword the building and growth of smart city.

The application of sensor networks in control and management of smart cities are increasing rapidly. The Internet of Things (IoT) platforms are becoming a major infrastructure component of the smart cities. The requirement for development of Web of Things (WoT) architectures and patterns have significant importance for optimal management and decision support in the smart city. Due to the fact that the sensor networks in even a small city produce a very large amount of data, manual interpretation of this data is not feasible. In recent years machine learning algorithms such as deep learning give the computer systems the ability to interpret and annotate this big data stream and to produce patterns which can help decision makers. This high-level interpretation is often presented in form of dashboards in the composition layers of the WoT. In this paper a deep learning based multisensor dashboard for decision support in the smart city is presented. The proposed dashboard is capable of interpretation and decision level fusion on sensor network data. Simulation results show that the proposed model is a suitable solution for making decisions in the IoT framework for the smart city.

This research navigates the confluence of data analytics, machine learning, and artificial intelligence to revolutionize the management of urban services in smart cities. The study thoroughly investigated with advanced tools to scrutinize key performance indicators integral to the functioning of smart cities, thereby enhancing leadership and decision-making strategies. Our work involves the implementation of various machine learning models such as Logistic Regression, Support Vector Machine, Decision Tree, Naive Bayes, and Artificial Neural Networks (ANN), to the data. Notably, the Support Vector Machine and Bernoulli Naive Bayes models exhibit robust performance with an accuracy rate of 70% precision score. In particular, the study underscores the employment of an ANN model on our existing dataset, optimized using the Adam optimizer. Although the model yields an overall accuracy of 61% and a precision score of 58%, implying correct predictions for the positive class 58% of the time, a comprehensive performance assessment using the Area Under the Receiver Operating Characteristic Curve (AUC-ROC) metrics was necessary. This evaluation results in a score of 0.475 at a threshold of 0.5, indicating that there's room for model enhancement. These models and their performance metrics serve as a key cog in our data analytics pipeline, providing decision-makers and city leaders with actionable insights that can steer urban service management decisions. Through real-time data availability and intuitive visualization dashboards, these leaders can promptly comprehend the current state of their services, pinpoint areas requiring improvement, and make informed decisions to bolster these services. This research illuminates the potential for data analytics, machine learning, and AI to significantly upgrade urban service management in smart cities, fostering sustainable and livable communities. Moreover, our findings contribute valuable knowledge to other cities aiming to adopt similar strategies, thus aiding the continued development of smart cities globally.

This chapter presents the Enhanced-real time Polling System (E-rtPS) with RT- Batch Markovian Arrival Process and proactive resource allocation framework to resolve the real-time network traffic issues and solve the interference problem optimistically. It proposes proactive resource allocation framework offers less computational overhead and yet again well-nigh performance in terms of resource allocation to maximize the transmission rate of each user. The chapter examines real-time results shows that the proposed framework outperforms the existing Institute of Electrical and Electronics Engineers (IEEE) 802.16 services in terms of throughput, session setup delay, and packet reception ratio. In IEEE 802.16, there are different Medium Access Control scheduling services, such as Unsolicited Grant Service, rtPS, and non-real-time Polling Service to provide better quality of services. The chapter presents an adaptive framework for dynamic speed management in smart cities. Since smart phone usage has risen to half a billion phones, mobile devices and connections account for 89 percent of the usage of mobile data traffic.

- Crowd density prediction is a critical task in modern public safety, urban planning, and large-scale event management. Accurately estimating the number of people and their spatial distribution enables authorities to prevent dangerous overcrowding and optimize resource deployment. This paper presents a hybrid crowd density prediction framework combining Convolutional Neural Networks (CNN) with Histogram of Oriented Gradients (HOG) feature descriptors. The CNN branch employs a modified VGG-16 backbone with dilated convolutions to capture high-level spatial representations, while the HOG branch extracts complementary gradient-based structural information from local image regions. Feature embeddings from both branches are fused through fully connected layers with dropout regularization to generate spatial density maps. The system is trained and validated on the ShanghaiTech Part A and Part B benchmark datasets. The proposed CNN+HOG hybrid achieves a Mean Absolute Error (MAE) of 75.4 and Root Mean Square Error (RMSE) of 112.3 on Part A, and MAE of 11.8, RMSE of 18.5 on Part B - outperforming pure CNN and HOG baselines. Training convergence analysis and error distribution studies confirm the robustness and generalization of the proposed approach across varying crowd densities and scene complexities. KeyWords: Crowd density prediction, convolutional neural network (CNN), histogram of oriented gradients (HOG), deep learning, density map estimation, ShanghaiTech dataset, feature fusion.

Counting the number of people in public places has received much attention, and researchers have devoted much effort to the task. However, the existing crowd counting approaches are mainly trained and tested in similar scenarios. The performance of crowd counting approaches degrades sharply when the test scenarios of the models are of different types from its training scenes. In practice, the crowd scenes are highly variable, and the lack of cross-scene capability could seriously limit the application of the existing approaches. We attribute the improvement in cross-scene crowd counting capability to the necessity of accommodating large changes in the scale of individuals and the ability to suppress the interference of cluttered backgrounds. To this end, we propose a multi-scale feature fusion network (MFFNet) with cascaded supervision. The multi-scale features extracted from the crowd images are upsampled and then combined into several feature blocks, followed by convolution and deconvolution operations on the feature blocks to derive feature matrices of different resolutions. The feature matrices are fused from bottom to top. In the process of feature fusion, the crowd density maps corresponding to the feature matrices of different resolutions are predicted separately. We devise cascaded supervision to synchronously optimize the network of different resolution density map prediction during training. The cross-scene crowd counting experiments are conducted on four types of scenes: ShanghaiTech Part_A (SHT A) with high-density crowd scenes and small-scale individuals, ShanghaiTech Part_B (SHT B) with sparse crowd distribution and medium-scale individuals, UCF_CC_50 dataset with extremely dense scenes and tiny scale individuals, and UCF-QNRF dataset with extreme variations. MFFNet exhibits the strongest scene adaptability relative to the state-of-the-art approaches, with an average decrease of 17.1% and 8.4% in mean absolute error (MAE) and mean square error (mse), respectively. The contributions of different components in our method are verified in the ablation study using the devised evaluation metrics. Our implementation will be available at <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/learnsharing/MFFNet</uri> .