Explore the vast range of titles available.

Cryoelectron tomography directly visualizes heterogeneous macromolecular structures in their native and complex cellular environments. However, existing computerassisted structure sorting approaches are low throughput or inherently limited due to their dependency on available templates and manual labels. Here, we introduce a high-throughput template-and-label-free deep learning approach, Deep Iterative Subtomogram Clustering Approach (DISCA), that automatically detects subsets of homogeneous structures by learning and modeling 3D structural features and their distributions. Evaluation on five experimental cryo-ET datasets shows that an unsupervised deep learning based method can detect diverse structures with a wide range of molecular sizes. This unsupervised detection paves the way for systematic unbiased recognition of macromolecular complexes in situ.

Hyperspectral image (HSI) clustering is a major challenge due to the redundant spectral information in HSIs. In this paper, we propose a novel deep subspace clustering method that extracts spatial–spectral features via contrastive learning. First, we construct positive and negative sample pairs through data augmentation. Then, the data pairs are projected into feature space using a CNN model. Contrastive learning is conducted by minimizing the distances of positive pairs and maximizing those of negative pairs. Finally, based on their features, spectral clustering is employed to obtain the final result. Experimental results gained over three HSI datasets demonstrate that our proposed method is superior to other state-of-the-art methods.

Hyperspectral image (HSI) clustering has drawn more and more attention in recent years as it frees us from labor-intensive manual annotation. However, current works cannot fully enjoy the rich spatial and spectral information due to redundant spectral signatures and fixed anchor learning. Moreover, the learned graph always obtains suboptimal results due to the separate affinity estimation and graph symmetry. To address the above challenges, in this paper, we propose large-scale hyperspectral image-projected clustering via doubly stochastic graph learning (HPCDL). Our HPCDL is a unified framework that learns a projected space to capture useful spectral information, simultaneously learning a pixel–anchor graph and an anchor–anchor graph. The doubly stochastic constraints are conducted to learn an anchor–anchor graph with strict probabilistic affinity, directly providing anchor cluster indicators via connectivity. Meanwhile, when using label propagation, pixel-level clustering results are obtained. An efficient optimization strategy is proposed to solve our HPCDL model, requiring monomial linear complexity concerning the number of pixels. Therefore, our HPCDL has the ability to deal with large-scale HSI datasets. Experiments on three datasets demonstrate the superiority of our HPCDL for both clustering performance and the time burden.

Tea leaves’ diseases caused by constant exposure to pathogens lead to significant crop yield loss globally. Diagnosing the tea leave disease at an early stage minimizes the tea yield loss. In this study, a novel approach is presented for automatically detecting tea leaves diseases based on image processing technology. The Non-dominated Sorting Genetic Algorithm (NSGA-II) based image clustering is proposed for detecting the disease area in tea leaves. After that, PCA and multi-class SVM is used for feature reduction and identifying the disease in the tea leaves, respectively. The result shows that the proposed algorithm can detect the type of disease persisting in tea leaves with an average accuracy of 83%. Five different tea leaf diseases are considered here, such as Red Rust, Red Spider, Thrips, Helopeltis, and Sunlight Scorching.

Digital photographs and visual data have become increasingly available, especially on the Web considered as the largest image database to date. However, the value of multimedia content depends on how easy it is to search and manage. Thus, the need to efficiently index, store, and retrieve images is becoming evermore important, particularly on the Web where existing image search and retrieval techniques do not seem to keep pace. Most existing solutions return a large quantity of search results ranked by their relevance to the user query. This can be tedious and time-consuming for the user, since the returned results usually contain multiple topics mixed together, and the user cannot be expected to have the time to scroll through the huge result list. A possible solution is to better organize the output information (prior or after query refinement), providing a means to facilitate the assimilation of the search results. In this context, image search result organization (ISRO) has been recently investigated as an effective and efficient solution to improve image retrieval quality on the Web. Most methods in this context exploit image clustering as a methodology capable of topic extraction and rendering semantically more meaningful results to the user. This survey paper provides a concise and comprehensive review of the methods related to cluster-based ISRO on the Web. It is made of four logical parts: First, we provide a glimpse on image information retrieval. Second, we briefly cover the background on ISRO. Third, we describe and categorize various steps involved in cluster-based ISRO, ranging over image representation, similarity computation, image clustering or grouping, and cluster-based search result visualization. Fourth, we briefly summarize and discuss ongoing research challenges and future directions, including high-dimensional feature indexing, joint word image modeling and implicit semantics, describing images based on aesthetics, automatic similarity metric learning, combining ensemble clustering methods, performing adaptive clustering, allowing dynamic trade-off between clustering quality and efficiency, diversifying image search results, integrating user feedback, and adapting results to mobile devices.

Brain tumour identification with traditional magnetic resonance imaging (MRI) tends to be time-consuming and in most cases, reading of the resulting images by human agents is prone to error, making it desirable to use automated image segmentation. This is a multi-step process involving: (a) collecting data in the form of raw processed or raw images, (b) removing bias by using pre-processing, (c) processing the image and locating the brain tumour, and (d) showing the tumour affected areas on a computer screen or projector. Several systems have been proposed for medical image segmentation but have not been evaluated in the field. This may be due to ongoing issues of image clarity, grey and white matter present in a scan image, lack of knowledge of the end user and constraints arising from MRI imaging systems. This makes it imperative to develop a comprehensive technique for the accurate diagnosis of brain tumors in MRI images. In this paper, we introduce a taxonomy consisting of ‘Data, Image segmentation processing, and View’ (DIV) which are the major components required to develop a high-end system for brain tumour diagnosis based on deep learning neural networks. The DIV taxonomy is evaluated based on system completeness and acceptance. The utility of the DIV taxonomy is demonstrated by classifying 30 state-of-the-art publications in the domain of medFical image segmentation systems based on deep neural networks. The results demonstrate that few components of medical image segmentation systems have been validated although several have been evaluated by identifying role and efficiency of the components in this domain.

Unsupervised hyperspectral image (HSI) clustering is a fundamental yet challenging task due to high dimensionality and complex spectral–spatial characteristics. In this paper, we propose a novel and efficient clustering framework centered on adaptive and diverse anchor graph modeling. First, we introduce a parameter-free construction strategy that employs Entropy Rate Superpixel (ERS) segmentation to generate multiple anchor graphs of varying sizes from a single HSI, overcoming the limitation of fixed anchor quantities and enhancing structural expressiveness. Second, we propose an anchor-to-pixel label propagation mechanism to transfer anchor-level cluster labels back to the pixel level, reinforcing spatial coherence and spectral discriminability. Third, we perform clustering directly at the anchor level, which substantially reduces computational cost while retaining structure-aware accuracy. Extensive experiments on three benchmark datasets (Trento, Salinas, and Pavia Center) demonstrate the effectiveness and efficiency of our approach.

Contrastive deep clustering has recently gained significant attention with its ability of joint contrastive learning and clustering via deep neural networks. Despite the rapid progress, previous works mostly require both positive and negative sample pairs for contrastive clustering, which rely on a relative large batch-size. Moreover, they typically adopt a two-stream architecture with two augmented views, which overlook the possibility and potential benefits of multi-stream architectures (especially with heterogeneous or hybrid networks). In light of this, this paper presents a new end-to-end deep clustering approach termed Heterogeneous Tri-stream Clustering Network (HTCN). The tri-stream architecture in HTCN consists of three main components, including two weight-sharing online networks and a target network, where the parameters of the target network are the exponential moving average of that of the online networks. Notably, the two online networks are trained by simultaneously (i) predicting the instance representations of the target network and (ii) enforcing the consistency between the cluster representations of the target network and that of the two online networks. Experimental results on four challenging image datasets demonstrate the superiority of HTCN over the state-of-the-art deep clustering approaches. The code is available at https://github.com/dengxiaozhi/HTCN .

Computational analysis is crucial to capitalize on the wealth of spatio-molecular information generated by mass spectrometry imaging (MSI) experiments. Currently, the spatial information available in MSI data is often under-utilized, due to the challenges of in-depth spatial pattern extraction. The advent of deep learning has greatly facilitated such complex spatial analysis. In this work, we use a pre-trained neural network to extract high-level features from ion images in MSI data, and test whether this improves downstream data analysis. The resulting neural network interpretation of ion images, coined neural ion images, is used to cluster ion images based on spatial expressions. We evaluate the impact of neural ion images on two ion image clustering pipelines, namely DBSCAN clustering, combined with UMAP-based dimensionality reduction, and k-means clustering. In both pipelines, we compare regular and neural ion images from two different MSI datasets. All tested pipelines could extract underlying spatial patterns, but the neural network-based pipelines provided better assignment of ion images, with more fine-grained clusters, and greater consistency in the spatial structures assigned to individual clusters. Additionally, we introduce the relative isotope ratio metric to quantitatively evaluate clustering quality. The resulting scores show that isotopical m/z values are more often clustered together in the neural network-based pipeline, indicating improved clustering outcomes. The usefulness of neural ion images extends beyond clustering towards a generic framework to incorporate spatial information into any MSI-focused machine learning pipeline, both supervised and unsupervised.

Deep clustering aims to boost clustering performance by learning powerful representations via deep learning. Despite their superiority over conventional shallow algorithms, autoencoder-based methods are typically hindered by heavy dependencies on large datasets and computationally expensive pre-training phases. Moreover, they often struggle to learn representations that are sufficiently discriminative for complex clustering tasks. To bridge this gap, we introduce a novel discriminative clustering framework utilizing Siamese encoders. By jointly training a Siamese encoder and a discriminative learning module, our method simultaneously captures robust features from data augmentations and imposes intra-cluster compactness. This dual optimization yields highly discriminative representations, which obviates the necessity for pre-training while ensuring rapid convergence and high accuracy. Extensive experiments on multiple benchmarks validate the superiority of our approach over state-of-the-art baselines.