Self-supervised learning (SSL) has garnered substantial interest within the machine learning and computer vision communities. Two prominent approaches in SSL include contrastive-based learning and self-distillation utilizing cropping augmentation. Lately, masked image modeling (MIM) has emerged as a more potent SSL technique, employing image inpainting as a pretext task. MIM creates a strong inductive bias toward meaningful spatial and semantic understanding. This has opened up new opportunities for SSL to contribute not only to classification tasks but also to more complex applications like object detection and image segmentation. Building upon this progress, our research paper introduces a scalable and practical SSL approach centered around more challenging pretext tasks that facilitate the acquisition of robust features. Specifically, we leverage multi-scale image reconstruction from randomly masked input images as the foundation for feature learning. Our hypothesis posits that reconstructing high-resolution images enables the model to attend to finer spatial details, particularly beneficial for discerning subtle intricacies within medical images. The proposed SSL features help improve classification performance on the Curated Breast Imaging Subset of Digital Database for Screening Mammography (CBIS-DDSM) dataset. In pathology classification, our method demonstrates a 3\% increase in average precision (AP) and a 1\% increase in the area under the receiver operating characteristic curve (AUC) when compared to state-of-the-art (SOTA) algorithms. Moreover, in mass margins classification, our approach achieves a 4\% increase in AP and a 2\% increase in AUC.
In this paper, we target the problem of fracture classification from clinical X-Ray images towards an automated Computer Aided Diagnosis (CAD) system. Although primarily dealing with an image classification problem, we argue that localizing the fracture in the image is crucial to make good class predictions. Therefore, we propose and thoroughly analyze several schemes for simultaneous fracture localization and classification. We show that using an auxiliary localization task, in general, improves the classification performance. Moreover, it is possible to avoid the need for additional localization annotations thanks to recent advancements in weakly-supervised deep learning approaches. Among such approaches, we investigate and adapt Spatial Transformers (ST), Self-Transfer Learning (STL), and localization from global pooling layers. We provide a detailed quantitative and qualitative validation on a dataset of 1347 femur fractures images and report high accuracy with regard to inter-expert correlation values reported in the literature. Our investigations show that i) lesion localization improves the classification outcome, ii) weakly-supervised methods improve baseline classification without any additional cost, iii) STL guides feature activations and boost performance. We plan to make both the dataset and code available.
In this paper, we propose a wide contextual residual network (WCRN) with active learning (AL) for remote sensing image (RSI) classification. Although ResNets have achieved great success in various applications (e.g. RSI classification), its performance is limited by the requirement of abundant labeled samples. As it is very difficult and expensive to obtain class labels in real world, we integrate the proposed WCRN with AL to improve its generalization by using the most informative training samples. Specifically, we first design a wide contextual residual network for RSI classification. We then integrate it with AL to achieve good machine generalization with limited number of training sampling. Experimental results on the University of Pavia and Flevoland datasets demonstrate that the proposed WCRN with AL can significantly reduce the needs of samples.
The focus of this paper is using a convolutional machine learning model with a modified U-Net structure for creating land cover classification mapping based on satellite imagery. The aim of the research is to train and test convolutional models for automatic land cover mapping and to assess their usability in increasing land cover mapping accuracy and change detection. To solve these tasks, authors prepared a dataset and trained machine learning models for land cover classification and semantic segmentation from satellite images. The results were analysed on three different land classification levels. BigEarthNet satellite image archive was selected for the research as one of two main datasets. This novel and recent dataset was published in 2019 and includes Sentinel-2 satellite photos from 10 European countries made in 2017 and 2018. As a second dataset the authors composed an original set containing a Sentinel-2 image and a CORINE land cover map of Estonia. The developed classification model shows a high overall F\textsubscript{1} score of 0.749 on multiclass land cover classification with 43 possible image labels. The model also highlights noisy data in the BigEarthNet dataset, where images seem to have incorrect labels. The segmentation models offer a solution for generating automatic land cover mappings based on Sentinel-2 satellite images and show a high IoU score for land cover classes such as forests, inland waters and arable land. The models show a capability of increasing the accuracy of existing land classification maps and in land cover change detection.
Performance of neural networks can be significantly improved by encoding known invariance for particular tasks. Many image classification tasks, such as those related to cellular imaging, exhibit invariance to rotation. We present a novel scheme using the magnitude response of the 2D-discrete-Fourier transform (2D-DFT) to encode rotational invariance in neural networks, along with a new, efficient convolutional scheme for encoding rotational equivariance throughout convolutional layers. We implemented this scheme for several image classification tasks and demonstrated improved performance, in terms of classification accuracy, time required to train the model, and robustness to hyperparameter selection, over a standard CNN and another state-of-the-art method.
The rapid development in additive manufacturing (AM), also known as 3D printing, has brought about potential risk and security issues along with significant benefits. In order to enhance the security level of the 3D printing process, the present research aims to detect and recognize illegal components using deep learning. In this work, we collected a dataset of 61,340 2D images (28x28 for each image) of 10 classes including guns and other non-gun objects, corresponding to the projection results of the original 3D models. To validate the dataset, we train a convolutional neural network (CNN) model for gun classification which can achieve 98.16% classification accuracy.
In recent years, the optical 4f system has drawn much attention in building high-speed and ultra-low-power optical neural networks (ONNs). Most optical systems suffer from the misalignment of the optical devices during installment. The performance of ONN based on the optical 4f system (4f-ONN) is considered sensitive to the misalignment in the optical path introduced. In order to comprehensively investigate the influence caused by the misalignment, we proposed a method for estimating the performance of a 4f-ONN in response to various misalignment in the context of the image classification task.The misalignment in numerical simulation is estimated by manipulating the optical intensity distributions in the fourth focus plane in the 4f system. Followed by a series of physical experiments to validate the simulation results. Using our method to test the impact of misalignment of 4f system on the classification accuracy of two popular image classification datasets, MNIST and Quickdraw16. On both datasets, we found that the performances of 4f-ONN generally degraded dramatically as the positioning error increased. Different positioning error tolerance in the misalignment orientations was observed over the two datasets. Classification performance could be preserved by positioning errors up to 200 microns in a specific direction.
The synthesis of high-resolution remote sensing images based on text descriptions has great potential in many practical application scenarios. Although deep neural networks have achieved great success in many important remote sensing tasks, generating realistic remote sensing images from text descriptions is still very difficult. To address this challenge, we propose a novel text-to-image modern Hopfield network (Txt2Img-MHN). The main idea of Txt2Img-MHN is to conduct hierarchical prototype learning on both text and image embeddings with modern Hopfield layers. Instead of directly learning concrete but highly diverse text-image joint feature representations for different semantics, Txt2Img-MHN aims to learn the most representative prototypes from text-image embeddings, achieving a coarse-to-fine learning strategy. These learned prototypes can then be utilized to represent more complex semantics in the text-to-image generation task. To better evaluate the realism and semantic consistency of the generated images, we further conduct zero-shot classification on real remote sensing data using the classification model trained on synthesized images. Despite its simplicity, we find that the overall accuracy in the zero-shot classification may serve as a good metric to evaluate the ability to generate an image from text. Extensive experiments on the benchmark remote sensing text-image dataset demonstrate that the proposed Txt2Img-MHN can generate more realistic remote sensing images than existing methods. Code and pre-trained models are available online (https://github.com/YonghaoXu/Txt2Img-MHN).
Data augmentation is widely used to enhance generalization in visual classification tasks. However, traditional methods struggle when source and target domains differ, as in domain adaptation, due to their inability to address domain gaps. This paper introduces GenMix, a generalizable prompt-guided generative data augmentation approach that enhances both in-domain and cross-domain image classification. Our technique leverages image editing to generate augmented images based on custom conditional prompts, designed specifically for each problem type. By blending portions of the input image with its edited generative counterpart and incorporating fractal patterns, our approach mitigates unrealistic images and label ambiguity, improving the performance and adversarial robustness of the resulting models. Efficacy of our method is established with extensive experiments on eight public datasets for general and fine-grained classification, in both in-domain and cross-domain settings. Additionally, we demonstrate performance improvements for self-supervised learning, learning with data scarcity, and adversarial robustness. As compared to the existing state-of-the-art methods, our technique achieves stronger performance across the board.
This paper presents one-bit supervision, a novel setting of learning from incomplete annotations, in the scenario of image classification. Instead of training a model upon the accurate label of each sample, our setting requires the model to query with a predicted label of each sample and learn from the answer whether the guess is correct. This provides one bit (yes or no) of information, and more importantly, annotating each sample becomes much easier than finding the accurate label from many candidate classes. There are two keys to training a model upon one-bit supervision: improving the guess accuracy and making use of incorrect guesses. For these purposes, we propose a multi-stage training paradigm which incorporates negative label suppression into an off-the-shelf semi-supervised learning algorithm. In three popular image classification benchmarks, our approach claims higher efficiency in utilizing the limited amount of annotations.
In multi-label classification, each example in a dataset may be annotated as belonging to one or more classes (or none of the classes). Example applications include image (or document) tagging where each possible tag either applies to a particular image (or document) or not. With many possible classes to consider, data annotators are likely to make errors when labeling such data in practice. Here we consider algorithms for finding mislabeled examples in multi-label classification datasets. We propose an extension of the Confident Learning framework to this setting, as well as a label quality score that ranks examples with label errors much higher than those which are correctly labeled. Both approaches can utilize any trained classifier. After demonstrating that our methodology empirically outperforms other algorithms for label error detection, we apply our approach to discover many label errors in the CelebA image tagging dataset.
The existing sonar image classification methods based on deep learning are often analyzed in Euclidean space, only considering the local image features. For this reason, this paper presents a sonar classification method based on improved Graph Attention Network (GAT), namely SI-GAT, which is applicable to multiple types imaging sonar. This method quantifies the correlation relationship between nodes based on the joint calculation of color proximity and spatial proximity that represent the sonar characteristics in non-Euclidean space, then the KNN (K-Nearest Neighbor) algorithm is used to determine the neighborhood range and adjacency matrix in the graph attention mechanism, which are jointly considered with the attention coefficient matrix to construct the key part of the SI-GAT. This SI-GAT is superior to several CNN (Convolutional Neural Network) methods based on Euclidean space through validation of real data.
Human action recognition has been a topic of interest across multiple fields ranging from security to entertainment systems. Tracking the motion and identifying the action being performed on a real time basis is necessary for critical security systems. In entertainment, especially gaming, the need for immediate responses for actions and gestures are paramount for the success of that system. We show that Motion History image has been a well established framework to capture the temporal and activity information in multi dimensional detail enabling various usecases including classification. We utilize MHI to produce sample data to train a classifier and demonstrate its effectiveness for action classification across six different activities in a single multi-action video. We analyze the classifier performance and identify usecases where MHI struggles to generate the appropriate activity image and discuss mechanisms and future work to overcome those limitations.
We propose an architecture for fine-grained visual categorization that approaches expert human performance in the classification of bird species. Our architecture first computes an estimate of the object's pose; this is used to compute local image features which are, in turn, used for classification. The features are computed by applying deep convolutional nets to image patches that are located and normalized by the pose. We perform an empirical study of a number of pose normalization schemes, including an investigation of higher order geometric warping functions. We propose a novel graph-based clustering algorithm for learning a compact pose normalization space. We perform a detailed investigation of state-of-the-art deep convolutional feature implementations and fine-tuning feature learning for fine-grained classification. We observe that a model that integrates lower-level feature layers with pose-normalized extraction routines and higher-level feature layers with unaligned image features works best. Our experiments advance state-of-the-art performance on bird species recognition, with a large improvement of correct classification rates over previous methods (75% vs. 55-65%).
Medical image classification models are frequently trained using training datasets derived from multiple data sources. While leveraging multiple data sources is crucial for achieving model generalization, it is important to acknowledge that the diverse nature of these sources inherently introduces unintended confounders and other challenges that can impact both model accuracy and transparency. A notable confounding factor in medical image classification, particularly in musculoskeletal image classification, is skeletal maturation-induced bone growth observed during adolescence. We train a deep learning model using a Covid-19 chest X-ray dataset and we showcase how this dataset can lead to spurious correlations due to unintended confounding regions. eXplanation Based Learning (XBL) is a deep learning approach that goes beyond interpretability by utilizing model explanations to interactively unlearn spurious correlations. This is achieved by integrating interactive user feedback, specifically feature annotations. In our study, we employed two non-demanding manual feedback mechanisms to implement an XBL-based approach for effectively eliminating these spurious correlations. Our results underscore the promising potential of XBL in constructing robust models even in the presence of confounding factors.
High spectral dimensionality and the shortage of annotations make hyperspectral image (HSI) classification a challenging problem. Recent studies suggest that convolutional neural networks can learn discriminative spatial features, which play a paramount role in HSI interpretation. However, most of these methods ignore the distinctive spectral-spatial characteristic of hyperspectral data. In addition, a large amount of unlabeled data remains an unexploited gold mine for efficient data use. Therefore, we proposed an integration of generative adversarial networks (GANs) and probabilistic graphical models for HSI classification. Specifically, we used a spectral-spatial generator and a discriminator to identify land cover categories of hyperspectral cubes. Moreover, to take advantage of a large amount of unlabeled data, we adopted a conditional random field to refine the preliminary classification results generated by GANs. Experimental results obtained using two commonly studied datasets demonstrate that the proposed framework achieved encouraging classification accuracy using a small number of data for training.
Medical image classification has developed rapidly under the impetus of the convolutional neural network (CNN). Due to the fixed size of the receptive field of the convolution kernel, it is difficult to capture the global features of medical images. Although the self-attention-based Transformer can model long-range dependencies, it has high computational complexity and lacks local inductive bias. Much research has demonstrated that global and local features are crucial for image classification. However, medical images have a lot of noisy, scattered features, intra-class variation, and inter-class similarities. This paper proposes a three-branch hierarchical multi-scale feature fusion network structure termed as HiFuse for medical image classification as a new method. It can fuse the advantages of Transformer and CNN from multi-scale hierarchies without destroying the respective modeling so as to improve the classification accuracy of various medical images. A parallel hierarchy of local and global feature blocks is designed to efficiently extract local features and global representations at various semantic scales, with the flexibility to model at different scales and linear computational complexity relevant to image size. Moreover, an adaptive hierarchical feature fusion block (HFF block) is designed to utilize the features obtained at different hierarchical levels comprehensively. The HFF block contains spatial attention, channel attention, residual inverted MLP, and shortcut to adaptively fuse semantic information between various scale features of each branch. The accuracy of our proposed model on the ISIC2018 dataset is 7.6% higher than baseline, 21.5% on the Covid-19 dataset, and 10.4% on the Kvasir dataset. Compared with other advanced models, the HiFuse model performs the best. Our code is open-source and available from https://github.com/huoxiangzuo/HiFuse.
Deep Neural Networks (DNNs) deliver state-of-the-art performance in many image recognition and understanding applications. However, despite their outstanding performance, these models are black-boxes and it is hard to understand how they make their decisions. Over the past few years, researchers have studied the problem of providing explanations of why DNNs predicted their results. However, existing techniques are either obtrusive, requiring changes in model training, or suffer from low output quality. In this paper, we present a novel method, NeuroMask, for generating an interpretable explanation of classification model results. When applied to image classification models, NeuroMask identifies the image parts that are most important to classifier results by applying a mask that hides/reveals different parts of the image, before feeding it back into the model. The mask values are tuned by minimizing a properly designed cost function that preserves the classification result and encourages producing an interpretable mask. Experiments using state-of-the-art Convolutional Neural Networks for image recognition on different datasets (CIFAR-10 and ImageNet) show that NeuroMask successfully localizes the parts of the input image which are most relevant to the DNN decision. By showing a visual quality comparison between NeuroMask explanations and those of other methods, we find NeuroMask to be both accurate and interpretable.
Zero-shot learning has been extensively investigated in the broader field of visual recognition, attracting significant interest recently. However, the current work on zero-shot learning in document image classification remains scarce. The existing studies either focus exclusively on zero-shot inference, or their evaluation does not align with the established criteria of zero-shot evaluation in the visual recognition domain. We provide a comprehensive document image classification analysis in Zero-Shot Learning (ZSL) and Generalized Zero-Shot Learning (GZSL) settings to address this gap. Our methodology and evaluation align with the established practices of this domain. Additionally, we propose zero-shot splits for the RVL-CDIP dataset. Furthermore, we introduce CICA (pronounced 'ki-ka'), a framework that enhances the zero-shot learning capabilities of CLIP. CICA consists of a novel 'content module' designed to leverage any generic document-related textual information. The discriminative features extracted by this module are aligned with CLIP's text and image features using a novel 'coupled-contrastive' loss. Our module improves CLIP's ZSL top-1 accuracy by 6.7% and GZSL harmonic mean by 24% on the RVL-CDIP dataset. Our module is lightweight and adds only 3.3% more parameters to CLIP. Our work sets the direction for future research in zero-shot document classification.
In recent advancement towards computer based diagnostics system, the classification of brain tumor images is a challenging task. This paper mainly focuses on elevating the classification accuracy of brain tumor images with transfer learning based deep neural network. The classification approach is started with the image augmentation operation including rotation, zoom, hori-zontal flip, width shift, height shift, and shear to increase the diversity in image datasets. Then the general features of the input brain tumor images are extracted based on a pre-trained transfer learning method comprised of Inception-v3. Fi-nally, the deep neural network with 4 customized layers is employed for classi-fying the brain tumors in most frequent brain tumor types as meningioma, glioma, and pituitary. The proposed model acquires an effective performance with an overall accuracy of 96.25% which is much improved than some existing multi-classification methods. Whereas, the fine-tuning of hyper-parameters and inclusion of customized DNN with the Inception-v3 model results in an im-provement of the classification accuracy.