Neural Architecture Search for Hyperspectral Image Classification: A Comprehensive Review and Future Perspectives

ETNAS: An energy consumption task-driven neural architecture search

Hyperspectral image (HSI) classification has become a core task in hyperspectral remote sensing interpretation, with deep learning dominating due to its ability to learn hierarchical features without manual engineering. As the model complexity has grown, manual design limitations have prompted a shift to automated approaches such as differentiable architecture search (DARTS), where the architectures are optimized for greater accuracy and efficiency. However, applying gradient-based neural architecture search (NAS) methods directly to hyperspectral classification presents several challenges. Regarding search space design, there is a lack of lightweight operators that can mitigate the spectral variability, spatial heterogeneity, and scale differences inherent in hyperspectral imagery. In terms of search strategy, the traditional DARTS approach directly derives the topology from operation weights, which can lead to suboptimal topological structures, and thus affects the performance of the network in HSI classification. In this article, to address these issues, we propose L3M, which is a lightweight multiscale and multiattention HSI classification network based on multistage search. The proposed approach introduces a novel lightweight operator to address the spectral variability, spatial heterogeneity, and scale differences in HSIs. The operation search and topology search are also decomposed into a multistage process to prevent a suboptimal network by searching for and determining the topological order of the candidate operations in a predefined operation space. L3M was validated on four public datasets, where the proposed model demonstrated a superior classification performance, compared to other lightweight models, while maintaining a low parameter count, low model complexity, and high inference speed.

Deep learning methods, especially deep convolutional neural networks (DCNNs), have been widely used in high-resolution remote sensing image (HRSI) semantic segmentation. In literature, most successful DCNNs are artificially designed through a large number of experiments, which often consume lots of time and depend on rich domain knowledge. Recently, neural architecture search (NAS), as a direction for automatically designing network architectures, has achieved great success in different kinds of computer vision tasks. For HRSI semantic segmentation, NAS faces two major challenges: (1) The task’s high complexity degree, which is caused by the pixel-by-pixel prediction demand in semantic segmentation, leads to a rapid expansion of the search space; (2) HRSI semantic segmentation often needs to exploit long-range dependency (i.e., a large spatial context), which means the NAS technique requires a lot of display memory in the optimization process and can be tough to converge. With the aforementioned considerations in mind, we propose a new decoupling NAS (DNAS) framework to automatically design the network architecture for HRSI semantic segmentation. In DNAS, a hierarchical search space with three levels is recommended: path-level, connection-level, and cell-level. To adapt to this hierarchical search space, we devised a new decoupling search optimization strategy to decrease the memory occupation. More specifically, the search optimization strategy consists of three stages: (1) a light super-net (i.e., the specific search space) in the path-level space is trained to get the optimal path coding; (2) we endowed the optimal path with various cross-layer connections and it is trained to obtain the connection coding; (3) the super-net, which is initialized by path coding and connection coding, is populated with kinds of concrete cell operators and the optimal cell operators are finally determined. It is worth noting that the well-designed search space can cover various network candidates and the optimization process can be done efficiently. Extensive experiments on the publicly open GID and FU datasets showed that our DNAS outperformed the state-of-the-art methods, including artificial networks and NAS methods.

Deep learning has made substantial breakthroughs in many fields due to its powerful automatic representation capabilities. It has been proven that neural architecture design is crucial to the feature representation of data and the final performance. However, the design of the neural architecture heavily relies on the researchers’ prior knowledge and experience. And due to the limitations of humans’ inherent knowledge, it is difficult for people to jump out of their original thinking paradigm and design an optimal model. Therefore, an intuitive idea would be to reduce human intervention as much as possible and let the algorithm automatically design the neural architecture.Neural Architecture Search(NAS) is just such a revolutionary algorithm, and the related research work is complicated and rich. Therefore, a comprehensive and systematic survey on the NAS is essential. Previously related surveys have begun to classify existing work mainly based on the key components of NAS: search space, search strategy, and evaluation strategy. While this classification method is more intuitive, it is difficult for readers to grasp the challenges and the landmark work involved. Therefore, in this survey, we provide a new perspective: beginning with an overview of the characteristics of the earliest NAS algorithms, summarizing the problems in these early NAS algorithms, and then providing solutions for subsequent related research work. In addition, we conduct a detailed and comprehensive analysis, comparison, and summary of these works. Finally, we provide some possible future research directions.

Neural Architecture Search (NAS) has emerged as a crucial method for automating the design of deep learning models. Despite its potential, NAS frequently requires substantial computational and hardware resources. To mitigate these challenges, transferable NAS (TNAS) has been introduced, leveraging prior NAS results to enhance performance on new tasks. However, existing methods largely focus on knowledge transfer within identical neural search spaces, overlooking the potential for cross-domain transferability. Motivated by this gap, we explore evolutionary TNAS across heterogeneous search spaces by learning common neural representations. In particular, we introduce a novel approach that encodes both operational and topological information of neural architectures into a unified sequence using a simple tokenizer. This sequence is then processed by a variational auto-encoder, with a Transformer-based encoder to capture rich neural representations and a decoder that reconstructs the original sequence. By utilizing these latent representations, we further establish an inter-domain mapping that acts as a bridge, enabling effective explicit solution transfer among diverse search spaces to enhance the evolutionary NAS process. To harness this capability, we develop an evolutionary sequential transfer optimization approach that transfers knowledge during population initialization, providing both flexibility and adaptability. To the best of our knowledge, this work serves as the first attempt in the literature exploring evolutionary TNAS across diverse spaces. Moreover, we demonstrate the utility of our method through comprehensive empirical studies using different architecture spaces, including NAS-Bench-101, NAS-Bench-201, and the DARTS search space. Our results show that the proposed method significantly enhances the adaptability and performance of NAS across varied domains.

A multi-perspective revisit to the optimization methods of Neural Architecture Search and Hyper-parameter optimization for non-federated and federated learning environments

In the future, vehicles will be equipped with increasingly advanced interactive intelligent electronic devices, which will induce drivers to conduct secondary tasks, thereby leading to distractions. Therefore, the detection and early warning of driver distraction are essential for improving driving safety and pose an important challenge in intelligent transportation systems. Previous studies used traditional machine learning and deep learning transfer models, which have the disadvantages of complicated and time-consuming manual feature engineering, strong subjectivity, and weak generalization performance. In this paper, we propose a fine-grained detection method for driver distraction based on neural architecture search. First, we design an automatic construction algorithm for deep convolutional neural networks based on neural architecture search, which automatically searches for the optimal deep convolutional neural network architecture without human involvement. In addition, we fuse driver-related multisource perception information, use an automatically constructed deep convolutional neural network to extract high-dimensional mapping features, and implement fine-grained detection of various types of driver distraction states. The results on a large-scale multimodal driver distraction dataset demonstrate that our method can efficiently search an optimal deep convolutional neural network, which can quickly converge, and can accurately detect the considered types of driver distraction states, the average detection accuracy reaches 99.7796%; moreover, it has satisfactory robustness.

Existing neural architecture search (NAS) methods comprise linear connected convolution operations and use ample search space to search task-driven convolution neural networks (CNN). These CNN models are computationally expensive and diminish the quality of receptive fields for tasks like micro-expression recognition (MER) with limited training samples. Therefore, we propose a refined neural architecture search strategy to search for a tiny CNN architecture for MER. In addition, we introduced a refined hybrid module (RHM) for inner-level search space and an optimal path explore network (OPEN) for outer-level search space. The RHM focuses on discovering optimal cell structures by incorporating a multilateral hybrid spatiotemporal operation space. Also, spatiotemporal attention blocks are embedded to refine the aggregated cell features. The OPEN search space aims to trace an optimal path between the cells to generate a tiny spatiotemporal CNN architecture instead of covering all possible tracks. The aggregate mix of RHM and OPEN search space availed the NAS method to robustly search and design an effective and efficient framework for MER. Compared with contemporary works, experiments reveal that the RNAS-MER is capable of bridging the gap between NAS algorithms and MER tasks. Furthermore, RNAS-MER achieves new state-of-the-art performances on challenging MER benchmarks, including 0.8511%, 0.7620%, 0.9078% and 0.8235% UAR on COMPOSITE, SMIC, CASME-II and SAMM datasets respectively.

Hyperspectral image (HSI) classification has been extensively applied for analyzing remotely sensed images. HSI data consist of multiple bands that provide abundant spatial information. Convolutional neural networks (CNNs) have emerged as powerful deep learning methods for processing visual data. In recent work, CNNs have shown impressive results in HSI classification. In this paper, we propose a hierarchical neural network architecture called feature extraction with hybrid spectral CNN (FE-HybridSN) to extract superior spectral–spatial features. FE-HybridSN effectively captures more spectral–spatial information while reducing computational complexity. Considering the prevalent issue of class imbalance in experimental datasets (IP, UP, SV) and real-world hyperspectral datasets, we apply the focal loss to mitigate these problems. The focal loss reconstructs the loss function and facilitates effective achievement of the aforementioned goals. We propose a framework (FEHN-FL) that combines FE-HybridSN and the focal loss for HSI classification and then conduct extensive HSI classification experiments using three remote sensing datasets: Indian Pines (IP), University of Pavia (UP), and Salinas Scene (SV). Using cross-entropy loss as a baseline, we assess the hyperspectral classification performance of various backbone networks and examine the influence of different spatial sizes on classification accuracy. After incorporating focal loss as our loss function, we not only compare the classification performance of the FE-HybridSN backbone network under different loss functions but also evaluate their convergence rates during training. The proposed classification framework demonstrates satisfactory performance compared to state-of-the-art end-to-end deep-learning-based methods, such as 2D-CNN, 3D-CNN, etc.

In recent years, deep-learning-based hyperspectral image (HSI) classification networks have become one of the most dominant implementations in HSI classification tasks. Among these networks, convolutional neural networks (CNNs) and attention-based networks have prevailed over other HSI classification networks. While convolutional neural networks with perceptual fields can effectively extract local features in the spatial dimension of HSI, they are poor at capturing the global and sequential features of spectral–spatial information; networks based on attention mechanisms, for example, Transformer, usually have better ability to capture global features, but are relatively weak in discriminating local features. This paper proposes a fusion network of convolution and Transformer for HSI classification, known as FusionNet, in which convolution and Transformer are fused in both serial and parallel mechanisms to achieve the full utilization of HSI features. Experimental results demonstrate that the proposed network has superior classification results compared to previous similar networks, and performs relatively well even on a small amount of training data.

High-resolution representations (HR) are essential for dense prediction tasks such as segmentation, detection, and pose estimation. Learning HR representations is typically ignored in previous Neural Architecture Search (NAS) methods that focus on image classification. This work proposes a novel NAS method, called HR-NAS, which is able to find efficient and accurate networks for different tasks, by effectively encoding multiscale contextual information while maintaining high-resolution representations. In HR-NAS, we renovate the NAS search space as well as its searching strategy. To better encode multiscale image contexts in the search space of HR-NAS, we first carefully design a lightweight transformer, whose computational complexity can be dynamically changed with respect to different objective functions and computation budgets. To maintain high-resolution representations of the learned networks, HR-NAS adopts a multi-branch architecture that provides convolutional encoding of multiple feature resolutions, inspired by HRNet [73]. Last, we proposed an efficient fine-grained search strategy to train HR-NAS, which effectively explores the search space, and finds optimal architectures given various tasks and computation resources. As shown in Fig. 1 (a), HR-NAS is capable of achieving state-of-the-art trade-offs between performance and FLOPs for three dense prediction tasks and an image classification task, given only small computational budgets. For example, HR-NAS surpasses SqueezeNAS [63] that is specially designed for semantic segmentation while improving efficiency by 45.9%. Code is available at https://github.com/dingmyu/HR-NAS.

There is a growing interest in automated neural architecture search (NAS) methods. They are employed to routinely deliver high-quality neural network architectures for various challenging data sets and reduce the designer's effort. The NAS methods utilizing multi-objective evolutionary algorithms are especially useful when the objective is not only to minimize the network error but also to minimize the number of parameters (weights) or power consumption of the inference phase. We propose a multi-objective NAS method based on Cartesian genetic programming for evolving convolutional neural networks (CNN). The method allows approximate operations to be used in CNNs to reduce the power consumption of a target hardware implementation. During the NAS process, a suitable CNN architecture is evolved together with approximate multipliers to deliver the best trade-offs between the accuracy, network size, and power consumption. The most suitable approximate multipliers are automatically selected from a library of approximate multipliers. Evolved CNNs are compared with common human-created CNNs of a similar complexity on the CIFAR-10 benchmark problem.

Convolutional neural networks (CNNs) have been shown to have the most advanced performance on image segmentation tasks. However, designing a reasonable convolutional neural network architecture requires a great deal of empirical knowledge, which can be expensive or unavailable in a variety of actual applications. To handle this problem, neural architecture search (NAS) has been developed by simultaneously optimizing the network architecture and hyperparameters, the main disadvantage of which is the requirement for extremely high computational resources. To accelerate the NAS of CNNs, a two-stage adaptive lightweight convolutional neural architecture search (AL-CNAS) method is proposed in this article based on an evolutionary algorithm. In this method, the original search space is divided into two parts based on feature grouping, and only one part is optimized at each stage to achieve an exponential decrease in the number of genotype combinations, which helps to improve search efficiency and reduce computational resource requirements. The proposed AL-CNAS is validated on the real-world segmentation of steel microstructure and two benchmark segmentation problems, and the computational results show that the proposed AL-CNAS has good segmentation performance and generality.

Deep learning techniques have boosted the performance of hyperspectral image (HSI) classification. In particular, convolutional neural networks (CNNs) have shown superior performance to that of the conventional machine learning algorithms. Recently, a novel type of neural networks called capsule networks (CapsNets) was presented to improve the most advanced CNNs. In this paper, we present a modified two-layer CapsNet with limited training samples for HSI classification, which is inspired by the comparability and simplicity of the shallower deep learning models. The presented CapsNet is trained using two real HSI datasets, i.e., the PaviaU (PU) and SalinasA datasets, representing complex and simple datasets, respectively, and which are used to investigate the robustness or representation of every model or classifier. In addition, a comparable paradigm of network architecture design has been proposed for the comparison of CNN and CapsNet. Experiments demonstrate that CapsNet shows better accuracy and convergence behavior for the complex data than the state-of-the-art CNN. For CapsNet using the PU dataset, the Kappa coefficient, overall accuracy, and average accuracy are 0.9456, 95.90%, and 96.27%, respectively, compared to the corresponding values yielded by CNN of 0.9345, 95.11%, and 95.63%. Moreover, we observed that CapsNet has much higher confidence for the predicted probabilities. Subsequently, this finding was analyzed and discussed with probability maps and uncertainty analysis. In terms of the existing literature, CapsNet provides promising results and explicit merits in comparison with CNN and two baseline classifiers, i.e., random forests (RFs) and support vector machines (SVMs).

Deep Learning and Beyond: AI’s New Horizons” is an expansive monograph that delves into the dynamic and rapidly evolving field of deep learning, a key area of artificial intelligence. This comprehensive work provides an insightful exploration of the revolutionary developments, foundational concepts, and future directions of deep learning, positioning itself as a crucial resource for understanding this transformative technology. The monograph commences with an introductory chapter that lays the groundwork by elucidating the background of machine learning and chronicling the ascent of deep learning. This foundational perspective is vital for comprehending the underlying principles and methodologies that have propelled deep learning to the forefront of technological innovation. Subsequent chapters delve into the “Historical Evolution” of AI, tracing the origins of neural networks from their early inception to the Renaissance period of deep learning. This historical analysis is pivotal in understanding the field’s trajectory and significant milestones. The “Foundations of Deep Learning” chapter presents an in-depth analysis of neural networks, covering their basic architecture, activation functions, layers, and the critical process of backpropagation and gradient descent. This is followed by examining “Advanced Neural Network Architectures,” where the focus shifts to state-of-the-art models such as convolutional neural networks, recurrent neural networks, and groundbreaking transformers and attention mechanisms. In addressing the “Training Deep Models,” the monograph explores regularization techniques, optimization strategies, and the increasingly significant role of transfer learning and pre-trained models. The application-oriented chapter, “Applications of Deep Learning,” showcases real-world implementations in domains like image recognition, natural language viii processing, and autonomous systems, illustrating deep learning technologies’ vast potential and versatility. The monograph also tackles the “Challenges in Deep Learning,” including overfitting, interpretability, and the ethical considerations surrounding AI deployment. It then ventures into the “Future of Deep Learning,” contemplating emerging trends such as integrating quantum computing, neural architecture search, and the convergence of neuroscience and deep learning. “Case Studies” provide practical insights and empirical evidence of the impact of deep learning. At the same time, the concluding chapter synthesizes the current state of deep learning, offers predictions for the future, and provides reflective thoughts on this transformative field. “Deep Learning and Beyond: AI’s New Horizons” educates and informs readers about the vast potential and future possibilities of deep learning and artificial intelligence. Keywords: Deep Learning, Artificial Intelligence, Neural Networks, Machine Learning, Convolutional Neural Networks, Recurrent Neural Networks, Transformers, Quantum Computing, Ethical AI, Neural Architecture Search, Natural Language Processing, Autonomous Systems, AI in Healthcare, Future of AI, AI Applications, AI Challenges.