Search Results (33)

Few-shot image classification requires models to generalize from limited labeled examples. While metric-based approaches such as Prototypical Networks have demonstrated strong performance, they rely exclusively on visual features and ignore the rich semantic information encoded in class names. This paper presents a systematic empirical study investigating the interaction between visual and semantic modalities in few-shot learning. We present Asymmetric Cross-Modal Prototypical Networks(ACM-ProtoNet), a controlled experimental framework which augments standard prototypical learning with frozen CLIP text encoders to incorporate zero-cost linguistic priors. Our method explicitly models the symmetric relationshipbetween visual and semantic modalities through learnable projection heads that map both image and text features into a shared embedding space. Image and text prototypes are fused via a learnable scalar gate $\alpha \in (0,1)$, allowing adaptive balancing of modalities. Under our experimental setup (frozen CLIP encoders, scalar fusion gate, simple template-based prompts), we observe an asymmetric pattern in comprehensive ablation studies on miniImageNet: cross-modal integration yields a statistically significant improvement in five-shot ($+2.12$ pp, $p=0.03125$, Wilcoxon signed-rank test over five seeds) but not in one-shot ($-0.09$ pp, n.s.) learning. Our key contribution is not achieving state-of-the-art accuracy but rather providing controlled empirical evidence about cross-modal interaction patterns under specific design constraints. Further analysis shows that: (1) structured semantic information is essential—random text features harm performance by $7.4$–$8.1$ percentage points; (2) projection heads provide asymmetric benefits, more critical in one-shot ($-2.85$ pp when removed) than in five-shot learning ($-0.74$ pp); (3) text-only prototypes achieve near-random performance (≈20%), suggesting that semantics alone are insufficient in our setup; (4) shuffled-class-name ablation confirms genuine semantic binding, where randomly permuting class-name assignments causes consistent degradation (five-shot: $-5.74$ pp, $p<0.001$; one-shot: $-3.83$ pp, $p<0.001$ across five seeds). These findings, specific to our simple fusion design, reveal an asymmetric pattern that is equally consistent with two hypotheses: (i) semantic priors may require sufficient visual context to be useful, or (ii) our scalar fusion gate may lack the capacity to leverage text in the extreme low-data regime of one-shot learning. This ambiguity motivates future work with more expressive fusion mechanisms and stronger text representations. Full article

Epileptic seizure (ES) detection is critical for improving clinical outcomes in epilepsy management. While intracranial EEG (iEEG) provides high-quality neural recordings, existing detection methods often rely on large amounts of data, involve high computational complexity, or fail to generalize in low-data settings. In this paper, we propose a lightweight, data-efficient, and high-performance approach for ES detection based on hyperdimensional computing (HDC). Our method first extracts local binary patterns (LBPs) from each iEEG channel to capture temporal–spatial dynamics. These binary sequences are then mapped into a high-dimensional space via HDC for robust representation, followed by a binary Naive Bayes classifier to distinguish ictal and inter-ictal states. The proposed design enables fast inference, low memory requirements, and suitability for hardware implementation. We evaluate the method on the SWEC-ETHZ iEEG short-term dataset. In one-shot learning, it achieves 100% sensitivity and specificity for most patients. In few-shot learning, it maintains 98.88% sensitivity and 93.09% specificity on average. The average latency is 4.31 s, demonstrating that it is much better than state-of-the-art methods. These results demonstrate the method’s potential for efficient, low-resource, and high-performance ES detection. Full article

This study introduces a physics-guided self-supervised framework for few-shot ultrasonic defect detection in concrete structures, addressing the dual challenges of scarce labels and domain variability in structural health monitoring (SHM). Our method integrates physics-informed augmentations, contrastive representation learning, and adversarial domain alignment within a mutually reinforcing cycle, enabling robust defect classification with minimal supervision. A Physics-Informed Augmentation Module synthesizes realistic ultrasonic signals, training a Transformer encoder to extract invariant features while suppressing sensor noise. An Adversarial Feature Aligner further improves cross-domain generalization by mitigating distribution shifts across heterogeneous concretes. Experimental validation on three benchmark datasets demonstrates 63–66% accuracy in one-shot cross-domain tasks and up to 89% in five-shot settings. These results represent 12–15 percentage point gains over modern few-shot baselines, with improvements statistically significant at p < 0.001. Compatible with existing ultrasonic hardware, the proposed framework bridges physics-based modeling and machine learning while paving the way for scalable, field-ready SHM solutions for aging infrastructure and resilient smart cities. Full article

Few-shot image classification aims to recognize novel classes from only a handful of labeled examples, a challenge in domains where data collection is costly or impractical. Existing solutions often rely on meta learning, fine tuning, or data augmentation, introducing computational overhead, risk of overfitting, or are not highly efficient. This paper introduces ProbaCLIP, a simple training-free approach that leverages Kernel Density Estimation (KDE) within the embedding space of Contrastive Language-Image Pre-training (CLIP). Unlike other CLIP-based methods, the proposed approach operates solely on visual embeddings and does not require text labels. Class-conditional probability densities were estimated from few-shot support examples, and queries were classified by likelihood evaluation, where Principal Component Analysis (PCA) was used for dimensionality reduction, compressing the dissimilarities between classes on each episode. We further introduced an optional bandwidth optimization strategy and a consensus decision mechanism through cross-validation, while addressing the special case of one-shot classification with distance-based measures. Extensive experiments on multiple datasets demonstrated that our method achieved competitive or superior accuracy compared to the state-of-the-art few-shot classifiers, reaching up to 98.37% accuracy in five-shot tasks and up to 99.80% in a 16-shot framework with ViT-L/14@336px. We proved our methodology by achieving high performance without gradient-based training, text supervision, or auxiliary meta-training datasets, emphasizing the effectiveness of combining pre-trained embeddings with statistical density estimation for data-scarce classification. Full article

To address the issue that prototypical networks in existing few-shot text classification methods suffer from performance limitations due to prototype shift and metric constraints, this paper proposes a meta-learning-based few-shot text classification method: Threshold Dynamic Multi-Source Decisive Prototypical Network (TDMP-Net) to solve these problems. This method designs two core components: the threshold dynamic data augmentation module and the multi-source information Decider. Specifically, the threshold dynamic data augmentation module achieves the optimization of the prototype estimation process by leveraging the multi-source information of query set samples, which thereby alleviates the prototype shift problem; meanwhile, the multi-source information Decider performs classification by relying on the multi-source information of the query set, thus alleviating the metric constraint problem. The effectiveness of the proposed method is verified on four benchmark datasets: under the five-way one-shot and five-way five-shot settings, TDMP-Net achieves average accuracies of 78.3% and 86.5%, respectively, which are an average improvement of 3.3 percentage points compared with current state-of-the-art methods. Experimental results show that this TDMP-Net can effectively alleviate the prototype shift problem and metric constraint problems, and has stronger generalization ability. Full article

Synthetic aperture radar (SAR) image classification under limited data conditions faces two major challenges: inter-class similarity, where distinct radar targets (e.g., tanks and armored trucks) have nearly identical scattering characteristics, and intra-class variability, caused by speckle noise, pose changes, and differences in depression angle. To address these challenges, we propose MHD-ProtoNet, a meta-learning framework that extends prototypical networks with two key innovations: margin-aware hard example mining to better separate confusable classes by enforcing prototype distance margins, and dual-loss optimization to refine embeddings and improve robustness to noise-induced variations. Evaluated on the MSTAR dataset in a five-way one-shot task, MHD-ProtoNet achieves $76.80\%$ accuracy, outperforming the Hybrid Inference Network (HIN) $(74.70\%)$, as well as standard few-shot methods such as prototypical networks $(69.38\%)$, ST-PN $(72.54\%)$, and graph-based models like ADMM-GCN $(61.79\%)$ and DGP-NET $(68.60\%)$. By explicitly mitigating inter-class ambiguity and intra-class noise, the proposed model enables robust SAR target recognition with minimal labeled data. Full article

Accurate and early detection of plant diseases is essential for effective management and the advancement of sustainable smart agriculture. However, building large annotated datasets for disease classification is often costly and time-consuming, requiring expert input. To address this challenge, this study explores the integration of few-shot learning with hyperspectral imaging to detect four major fungal diseases in tomato plants: Alternaria alternata, Alternaria solani, Botrytis cinerea, and Fusarium oxysporum. Following inoculation, hyperspectral images were captured every other day from Day 1 to Day 7 post inoculation. The proposed hybrid method includes three main steps: (1) preprocessing of hyperspectral image cubes, (2) deep feature extraction using the EfficientNet model, and (3) classification using Manhattan distance within a few-shot learning framework. This combination leverages the strengths of both spectral imaging and deep learning for robust detection with minimal data. The few-shot learning approach achieved high detection accuracies of 85.73%, 80.05%, 90.33%, and 82.09% for A. alternata, A. solani, B. cinerea, and F. oxysporum, respectively, based on data collected on Day 7 post inoculation using only three training images per class. Accuracy improved over time, reflecting the progressive nature of symptom development and the model’s adaptability with limited data. Notably, A. alternata and B. cinerea were reliably detected by Day 3, while A. solani and F. oxysporum reached dependable detection levels by Day 5. Routine visual assessments showed that A. alternata and B. cinerea developed visible symptoms by Day 5, whereas A. solani and F. oxysporum remained asymptomatic until Day 7. The model’s ability to detect infections up to two days before visual symptoms emerged highlights its value for pre-symptomatic diagnosis. These findings support the use of few-shot learning and hyperspectral imaging for early, accurate disease detection, offering a practical solution for precision agriculture and timely intervention. Full article

Generalized few-shot semantic segmentation (GFSS), which requires strong segmentation performance on novel classes while retaining the performance on base classes, is attracting increasing attention. Recent studies have demonstrated the effectiveness of applying visual prompts to solve GFSS problems, but there are still unresolved issues. Due to the confusion between the backgrounds and novel classes foreground during base class pre-training, the learned base visual prompts will mislead the novel visual prompts during novel class fine-tuning, leading to sub-optimal results. This paper proposes a background-enhanced visual prompting Transformer (Beh-VPT) to solve the problem. Specifically, we innovatively propose background visual prompts, which can learn potential novel class information in the background during base class pre-training and transfer the information to novel visual prompts during novel class fine-tuning via our proposed Hybrid Causal Attention Module. Additionally, we propose a background-enhanced segmentation head that is used in conjunction with background prompts to enhance the model’s capacity for learning novel classes. Considering the GFSS settings that take into account both base and novel classes, we introduce Singular Value Fine-Tuning in the non-meta learning paradigm to further unleash the full potential of the model. Extensive experiments show that the proposed method achieves state-of-the-art performance for GFSS on PASCAL-$5^{\mathrm{i}}$ and COCO-${20}^{\mathrm{i}}$ datasets. For example, considering both base and novel classes, the improvements in mIoU range from 0.47% to 1.08% (COCO-${20}^{\mathrm{i}}$) in the one-shot and five-shot scenarios, respectively. In addition, our method does not cause a fallback of mIoU in base classes relative to the baseline. Full article

As demand for high-capacity, low-latency communication rises, mmWave systems are essential for enabling ultra-high-speed transmission in fifth-generation mobile communication technology (5G) and upcoming 6G networks, especially in dynamic, data-scarce environments. However, deploying mmWave systems in dynamic environments presents significant challenges, especially in beam selection, where limited training data and environmental variability hinder optimal performance. In such scenarios, computation offloading has emerged as a key enabler, allowing computationally intensive tasks to be shifted from resource-constrained edge devices to powerful cloud servers, thereby reducing latency and optimizing resource utilization. This paper introduces a novel cloud–edge collaborative approach integrating few-shot learning (FSL) with multimodal fusion to address these challenges. By leveraging data from diverse modalities—such as red-green-blue (RGB) images, radar signals, and light detection and ranging (LiDAR)—within a cloud–edge architecture, the proposed framework effectively captures spatiotemporal features, enabling efficient and accurate beam selection with minimal data requirements. The cloud server is tasked with computationally intensive training, while the edge node focuses on real-time inference, ensuring low-latency decision making. Experimental evaluations confirm the model’s robustness, achieving high beam selection accuracy under one-shot and five-shot conditions while reducing computational overhead. This study highlights the potential of combining cloud–edge collaboration with FSL and multimodal fusion for next-generation wireless networks, paving the way for scalable, intelligent, and adaptive mmWave communication systems. Full article

In recent years, the fields of one-shot and few-shot object detection and classification have garnered significant attention. However, the rapid adaptation of robots to previously unencountered or novel environments remains a formidable challenge. Inspired by biological learning processes, meta-learning seeks to replicate the way humans and animals quickly adapt to new tasks by leveraging prior knowledge and generalizing across experiences. Despite this, traditional meta-learning methods that rely on deepening or widening neural networks offer only marginal improvements in model performance. To address this, we proposed a novel framework termed Residual Multi-Attention Domain-Adaptive Meta-Learning (DRMA-DAML). Our framework, motivated by biological principles like the human visual system’s concurrent handling of global and local details for enhanced perception and decision making, empowers the model to significantly enhance performance without augmenting the depth of the neural network, thus avoiding the overfitting and vanishing gradient problems typical of deeper architectures. Empirical evidence from both simulated environments and real-world applications demonstrates that DRMA-DAML achieves state-of-the-art performance. Specifically, it improves adaptation accuracy by 11.18% on benchmark tasks and achieves a 97.64% success rate in real-world object manipulation, surpassing existing methods. These results validate the effectiveness of our approach in rapid adaptation for robotic systems. Full article

Urban cycling is a sustainable mode of transportation in large cities, and it offers many advantages. It is an eco-friendly means of transport that is accessible to the population and easy to use. Additionally, it is more economical than other means of transportation. Urban cycling is beneficial for physical health and mental well-being. Achieving sustainable mobility and the evolution towards smart cities demands a comprehensive analysis of all the essential aspects that enable their inclusion. Road safety is particularly important, which must be prioritized to ensure safe transportation and reduce the incidence of road accidents. In order to help reduce the number of accidents that urban cyclists are involved in, this work proposes an alternative solution in the form of an intelligent computational assistant that utilizes simplified machine learning (SML) to detect potential risks of unexpected collisions. This technological approach serves as a helpful alternative to the current problem. Through our methodology, we were able to identify the problem involved in the research, design, and development of the solution proposal; collect and analyze data; and obtain preliminary results. These results experimentally demonstrate how the proposed model outperforms most state-of-the-art models that use a metric learning layer for small image sets. Full article

Early detection of plant diseases is crucial for agro-holdings, farmers, and smallholders. Various neural network architectures and training methods have been employed to identify optimal solutions for plant disease classification. However, research applying one-shot or few-shot learning approaches, based on similarity determination, to the plantdisease classification domain remains limited. This study evaluates different loss functions used in similarity learning, including Contrastive, Triplet, Quadruplet, SphereFace, CosFace, and ArcFace, alongside various backbone networks, such as MobileNet, EfficientNet, ConvNeXt, and ResNeXt. Custom datasets of real-life images, comprising over 4000 samples across 68 classes of plant diseases, pests, and their effects, were utilized. The experiments evaluate standard transfer learning approaches alongside similarity learning methods based on two classes of loss function. Results demonstrate the superiority of cosine-based methods over Siamese networks in embedding extraction for disease classification. Effective approaches for model organization and training are determined. Additionally, the impact of data normalization is tested, and the generalization ability of the models is assessed using a special dataset consisting of 400 images of difficult-to-identify plant disease cases. Full article

In this research, inspired by the principles of biological visual attention mechanisms and swarm intelligence found in nature, we present an Enhanced Self-Correlation Attention and Multi-Branch Joint Module Network (EMNet), a novel model for few-shot image classification. Few-shot image classification aims to address the problem of image classification when data are limited. Traditional models require a large amount of labeled data for training, while few-shot learning trains models using only a small number of samples (just a few samples per class) to recognize new categories. EMNet shows its potential for bio-inspired algorithms in optimizing feature extraction and enhancing generalization capabilities. It features two key modules: Enhanced Self-Correlated Attention (ESCA) and Multi-Branch Joint Module (MBJ Module). EMNet tackles two main challenges in few-shot learning: how to make an effective important feature extraction and enhancement in images, and improving generalization to new categories. The ESCA module boosts the precision in extracting crucial local features, enhancing classification accuracy. The MBJ module focuses on shared features across images, emphasizing similarities within classes and subtle differences between them. This enhances model adaptability and generalization to new categories. Experimental results show that our model performs better than existing models in one-shot and five-shot tasks on mini-ImageNet, CUB-200, and CIFAR-FS datasets, which proves the proposed model to be an efficient end-to-end solution for few-shot image classification. In the five-way one-shot and five-way five-shot experiments on the CUB-200-2011 dataset, EMNet achieved classification accuracies that were 1.27 and 0.54 percentage points higher than those of RENet, respectively. In the five-way one-shot and five-way five-shot experiments on the miniImageNet dataset, EMNet’s classification accuracies were 0.02 and 0.48 percentage points higher than those of RENet, respectively. In the five-way one-shot and five-way five-shot experiments on the CIFAR-FS dataset, EMNet’s classification accuracies were 0.19 and 0.18 percentage points higher than those of RENet. Full article

Tomato fungal diseases can cause significant economic losses to farmers. Advanced disease detection methods based on symptom recognition in images face challenges when identifying fungal diseases in tomatoes, especially with limited training images. This study utilized novel techniques designed for limited data scenarios, such as one-shot and few-shot learning, to identify three tomato fungal diseases, i.e., Alternaria solani, Alternaria alternata, and Botrytis cinerea. Automated feature extraction was performed using the ResNet-12 deep model, and a cosine similarity approach was employed during shot learning. The accuracy of diagnosing the three diseases and healthy leaves using the 4-way 1-shot learning method was 91.64, 92.37, 92.93, and 100%. For the 4-way 3-shot learning method, the accuracy improved to 92.75, 95.07, 96.63, and 100%, respectively. These results demonstrate that the proposed method effectively reduces the dependence on experts labeling images, working well with small datasets and enhancing plant disease identification. Full article

This paper explores the use of few-shot learning in Wi-Fi-based indoor positioning, utilizing convolutional neural networks (CNNs) combined with meta-learning techniques to enhance the accuracy and efficiency of positioning systems. The focus is on addressing the challenge of limited labeled data, a prevalent issue in extensive indoor environments. The study explores various scenarios, comparing the performance of the base CNN and meta-learning models. The meta-learning approach involves few-shot learning tasks, such as three-way N-shot, five-way N-shot, etc., to enhance the model’s ability to generalize from limited data. The experiments were conducted across various scenarios, evaluating the performance of the models with different numbers of samples per class (K) after filtering by cosine similarity (FCS) during both the stages of data preprocessing and meta-learning. The scenarios included both base classes and novel classes, with and without meta-learning. The results indicated that the base CNN model achieved varying accuracy levels depending on the scenario and the number of samples per class retained after FCS. Meta-learning performed acceptably in scenarios with fewer samples, which are the distinct datasets pertaining to novel classes. With 20 samples per class, the base CNN achieved an accuracy of 0.80 during the pre-training stage, while meta-learning (three-way one-shot) achieved an accuracy of 0.78 on a new small dataset with novel classes. Full article