Accurate classification of lung adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC), the two major subtypes of non-small cell lung cancer (NSCLC), remains challenging due to heterogeneous imaging characteristics and overlapping radiological-pathological patterns. To address this, we propose M2HCNet, a Multimodal and Multiscale Hierarchical Classification Network that integrates computed tomography (CT) and histopathology images for robust lung cancer subtype identification. A multi-scale feature extraction module captures both fine-grained cellular details and global contextual structures, enabling complementary fusion across modalities. The extracted features are then aligned using an optimal transport-based contrastive alignment strategy, which reduces distributional discrepancies and enhances cross-modal consistency under weakly paired conditions. A hierarchical classification mechanism further models inter-class relationships and coarse-to-fine label dependencies to improve discrimination between closely related subtypes. Experiments conducted on the TCGA-LUAD/LUSC cohort demonstrate that M2HCNet achieves 93.8% accuracy, 91.9% F1-score, and 95.5% AUC, outperforming ten recent state-of-the-art baselines. Ablation studies confirm the contributions of multi-scale fusion, cross-modal alignment, and hierarchical learning to overall performance. These findings suggest that M2HCNet effectively leverages heterogeneous yet complementary modalities for reliable lung cancer subtype classification. Future work will extend the framework to patient-matched and multi-institutional datasets to enhance generalization and clinical applicability.

Metric learning is n fundamental problem in computer vision. Different features and algorithms may tackle a problem from different angles, and thus often provide complementary information. In this paper; we propose a fusion algorithm which outputs enhanced metrics by combining multiple given metrics (similarity measures). Unlike traditional co-training style algorithms where multi-view features or multiple data subsets are used for classification or regression, we focus on fusing multiple given metrics through diffusion process in an unsupervised way. Our algorithm has its particular advantage when the input similarity' matrices are the outputs from diverse algorithms. We provide both theoretical and empirical explanations to our method. Significant improvements over the state-of-the-art results have been observed on various benchmark datasets. For example, we have achieved 100% accuracy (no longer the bull's eye measure) on the MPEG-7 shape dataset. Our method has a wide range of applications in machine learning and computer vision.

This paper extends the neighborhood components analysis method (NCA) to learning a mixture of sparse distance metrics for classification and dimensionality reduction. We emphasize two important properties in the recent learning literature, locality and sparsity, and (1) pursue a set of local distance metrics by maximizing a conditional likelihood of observed data; and (2) add l1-norm of eigenvalues of the distance metric to favor low rank matrices of fewer parameters. Experimental results on standard UCI machine learning datasets, face recognition datasets, and image categorization datasets demonstrate the feasibility of our approach for both distance metric learning and dimensionality reduction.

In this paper, we propose a diffusion-based approach to improve an input similarity metric. The diffusion process propagates similarity mass along the intrinsic manifold of data points. Our approach results in a global similarity metric which differs from the query-specific one for ranking produced by label propagation [26]. Unlike diffusion maps [7], our approach directly improves a given similarity metric without introducing any extra distance notions. We call our approach Self-Smoothing Operator (SSO). To demonstrate its wide applicability, experiments are reported on image retrieval, clustering, classification, and segmentation tasks. In most cases, using SSO results in significant performance gains over the original similarity metrics, with also very evident advantage over diffusion maps.

The Conditional Random Fields (CRF) model, using patch-based classification bound with context information, has recently been widely adopted for image segmentation/ labeling. In this paper, we propose three components for improving the speed and accuracy, and illustrate them on a recently developed auto-context algorithm [28]: (1) a new coding scheme for multiclass classification, named data-assisted output code (DAOC); (2) a scale-space approach to make it less sensitive to geometric scale change; and (3) a region-based voting scheme to make it faster and more accurate at object boundaries. The proposed multiclass classifier, DAOC, is general and particularly appealing when the number of class becomes large since it needs a minimal number of [log2k] binary classifiers for k classes. We show advantages of the DAOC classifier over the existing algorithms on several Irvine repository datasets, as well as vision applications. Combining DAOC, the scale-space approach, and the region-based voting scheme for autocontext, the overall algorithm is significantly faster (5 ∼ 10 times) than the original auto-context, with improved accuracy over many of the existing algorithms on theMSRC [24] and VOC 2007 [7] datasets.

This paper describes a coarse-to-fine learning based image registration algorithm which has particular advantages in dealing with multi-modality images. Many existing image registration algorithms [18] use a few designed terms or mutual information to measure the similarity between image pairs. Instead, we push the learning aspect by selecting and fusing a large number of features for measuring the similarity. Moreover, the similarity measure is carried in a coarse-to-fine strategy: global similarity measure is first performed to roughly locate the component, we then learn/compute similarity on the local image patches to capture the fine level information. When estimating the transformation parameters, we also engage a coarse-to-fine strategy. Off-the-shelf interest point detectors such as SIFT [12] have degraded results on medical images. We further push the learning idea to extract the main structures/landmarks. Our algorithm is illustrated on three applications: (1) registration of mouse brain images of different modalities, (2) registering human brain image of MRI T1 and T2 images, (3) faces of different expressions. We show greatly improved results over the existing algorithms based on either mutual information or geometric structures.

Manifold learning algorithms have been recently reported superior to classical dimensionality reduction techniques, such as PCA or MDS, in their ability to discover a more meaningful low-dimensional embedding of the high-dimensional samples. However, most of them encounter the problem of extension to novel samples. In this paper, we propose a regression model to extend three well-known manifold learning algorithms, i.e. Isomap, LLE, and Laplacian Eigenmap to novel samples by neural networks. We first examine these algorithms, and then show that the nonlinear dimensionality reduction ability can be acquired by neural networks, thus the extension problem is easily addressed. This model is very flexible and still preserves the nonlinear nature of the manifold leaning algorithms. Experimental results of data visualization and classification are reported, which validate the feasibility of the proposed model.