Semi-supervised Image Classification Research Articles

Leveraging Frame- and Feature-Level Progressive Augmentation for Semi-supervised Action Recognition

Semi-supervised action recognition is a challenging yet prospective task due to its low reliance on costly labeled videos. One high-profile solution is to explore frame-level weak/strong augmentations for learning abundant representations, inspired by the FixMatch framework dominating the semi-supervised image classification task. However, such a solution mainly brings perturbations in terms of texture and scale, leading to the limitation in learning action representations in videos with spatiotemporal redundancy and complexity. Therefore, we revisit the creative trick of weak/strong augmentations in FixMatch, and then propose a novel Frame- and Feature-level augmentation FixMatch (dubbed as F 2 -FixMatch) framework to learn more abundant action representations for being robust to complex and dynamic video scenarios. Specifically, we design a new Progressive Augmentation (P-Aug) mechanism that implements the weak/strong augmentations first at the frame level, and further implements the perturbation at the feature level, to obtain abundant four types of augmented features in broader perturbation spaces. Moreover, we present an evolved Multihead Pseudo-Labeling (MPL) scheme to promote the consistency of features across different augmented versions based on the pseudo labels. We conduct extensive experiments on several public datasets to demonstrate that our F 2 -FixMatch achieves the performance gain compared with current state-of-the-art methods. The source codes of F 2 -FixMatch are publicly available at https://github.com/zwtu/F2FixMatch.

Discriminative and robust least squares regression for semi-supervised image classification

Due to the ability to leverage information from both unlabeled and labeled data, semi-supervised classification has found extensive applications in various practical scenarios. However, there are two major drawbacks: (1) Traditional graph construction leads to high algorithmic complexity, which limits efficiency. (2) The decision boundary might be blurred by boundary points in common cases. To cope with these issues, inspired by classical Least Squares Regression (LSR), we present a novel semi-supervised classification algorithm termed as Discriminative and Robust LSR (DRLSR) for semi-supervised image classification. First of all, a manifold regularization term is designed and introduced to an LSR-based semi-supervised method to preserve local manifold structures and smooth structures of the subspace, which strengthens the discrimination ability. Meanwhile, preservation of local manifold structures also contributes to restrain decision boundaries from being blurred by boundary points, which strengthens robustness of the algorithm. After that, an efficient alternative optimization method is applied to our algorithm. Evidence of the effectiveness of DRLSR are compelled by extensive experimental results.

De-biased Teacher: Rethinking IoU Matching for Semi-supervised Object Detection

Most of the recent research in semi-supervised object detection follows the pseudo-labeling paradigm evolved from the semi-supervised image classification task. However, the training paradigm of the two-stage object detector inevitably makes the pseudo-label learning process for unlabeled images full of bias. Specifically, the IoU matching scheme used for selecting and labeling candidate boxes is based on the assumption that the matching source~(ground truth) is accurate enough in terms of the number of objects, object position and object category. Obviously, pseudo-labels generated for unlabeled images cannot satisfy such a strong assumption, which makes the produced training proposals extremely unreliable and thus severely spoil the follow-up training. To de-bias the training proposals generated by the pseudo-label-based IoU matching, we propose a general framework -- De-biased Teacher, which abandons both the IoU matching and pseudo labeling processes by directly generating favorable training proposals for consistency regularization between the weak/strong augmented image pairs. Moreover, a distribution-based refinement scheme is designed to eliminate the scattered class predictions of significantly low values for higher efficiency. Extensive experiments demonstrate that the proposed De-biased Teacher consistently outperforms other state-of-the-art methods on the MS-COCO and PASCAL VOC benchmarks. Source codes are available at https://github.com/wkfdb/De-biased-Teracher.

Semi-Supervised Land Cover Classification of Remote Sensing Imagery Using CycleGAN and EfficientNet

Image classification of very high resolution (VHR) images is a fundamental task in the remote sensing domain for various applications, such as land cover mapping, vegetation mapping, and urban planning. Recently, deep learning-based semantic segmentation networks demonstrated the promising performance for pixel-level image classification. However, deep learning-based approaches are generally limited by the requirement of a sufficient amount of labeled data to obtain stable accuracy, and acquiring reference labels of remotely-sensed VHR images is very labor-extensive and expensive. Hence, this paper applied a semi-supervised learning-based CycleGAN and EfficientNet for VHR remote sensing image classification to overcome this problem. The proposed method achieved the highest accuracy than the other benchmarks. The largest increase in accuracy was observed in a test site containing complex objects due to the regularization effect of the semi-supervised method using unlabeled data. Moreover, results indicated that a relatively sufficient amount of unlabeled data compared with labeled data are required to increase the classification accuracy by controlling the amount of labeled and unlabeled data. Finally, we verified that the semi-supervised method returned significantly improved results irrespective of the three classification network structures, displaying the applicability of the method for semi-supervised image classification on remotely-sensed VHR images.

Top-K Pseudo Labeling for Semi-Supervised Image Classification

In this paper, a top-k pseudo labeling method for semi-supervised self-learning is proposed. Pseudo labeling is a key technology in semi-supervised self-learning. Briefly, the quality of the pseudo label generated largely determined the convergence of the neural network and the accuracy obtained. In this paper, the authors use a method called top-k pseudo labeling to generate pseudo label during the training of semi-supervised neural network model. The proposed labeling method helps a lot in learning features from unlabeled data. The proposed method is easy to implement and only relies on the neural network prediction and hyper-parameter k. The experiment results show that the proposed method works well with semi-supervised learning on CIFAR-10 and CIFAR-100 datasets. Also, a variant of top-k labeling for supervised learning named top-k regulation is proposed. The experiment results show that various models can achieve higher accuracy on test set when trained with top-k regulation.

Graph Convolutional Networks based on manifold learning for semi-supervised image classification

Due to a huge volume of information in many domains, the need for classification methods is imperious. In spite of many advances, most of the approaches require a large amount of labeled data, which is often not available, due to costs and difficulties of manual labeling processes. In this scenario, unsupervised and semi-supervised approaches have been gaining increasing attention. The GCNs (Graph Convolutional Neural Networks) represent a promising solution since they encode the neighborhood information and have achieved state-of-the-art results on scenarios with limited labeled data. However, since GCNs require graph-structured data, their use for semi-supervised image classification is still scarce in the literature. In this work, we propose a novel approach, the Manifold-GCN, based on GCNs for semi-supervised image classification. The main hypothesis of this paper is that the use of manifold learning to model the graph structure can further improve the GCN classification. To the best of our knowledge, this is the first framework that allows the combination of GCNs with different types of manifold learning approaches for image classification. All manifold learning algorithms employed are completely unsupervised, which is especially useful for scenarios where the availability of labeled data is a concern. A broad experimental evaluation was conducted considering 5 GCN models, 3 manifold learning approaches, 3 image datasets, and 5 deep features. The results reveal that our approach presents better accuracy than traditional and recent state-of-the-art methods with very efficient run times for both training and testing.

Improving Semi-Supervised Image Classification by Assigning Different Weights to Correctly and Incorrectly Classified Samples

Semi-supervised deep learning, a model that aims to effectively use unlabeled data to help learn sample features from labeled data, is a recent hot topic. To effectively use unlabeled data, a new semi-supervised learning model based on a consistency strategy is proposed. In the supervised part with labeled samples, the image generation model first generates some artificial images to complement the limited number of labeled samples. Secondly, the sample label mapping, as the “benchmark”, is compared to the corresponding sample features in the network as an additional loss to complement the original supervisory loss, aiming to better correct the model parameters. Finally, the original supervised loss is changed so that the network parameters are determined by the characteristics of each correctly classified sample. In the unsupervised part, the actual unsupervised loss is altered so that the model does not “treat all samples equally” and can focus more on the characteristics of misclassified samples. A total of 40 labeled samples from the CIFAR-10 and SVHN datasets were used to train the semi-supervised model achieving accuracies of 93.25% and 96.83%, respectively, demonstrating the effectiveness of the proposed semi-supervised model.

Balanced and Accurate Pseudo-Labels for Semi-Supervised Image Classification

Image classification by semi-supervised learning has recently become a hot spot, and the Co-Training framework is an important method of semi-supervised image classification. In the traditional Co-Training structure, the sub-networks will generate pseudo-labels for each other, and these pseudo-labels will further be used as a supervisory signal for model training. However, the pseudo-labels will hurt classification performance because of their low accuracy and unbalanced distribution. In this article, we are trying to solve the preceding two problems by designing the Balanced Module (BM) and Gaussian Mixture Module (GMM), and propose BAPS (the B alanced and A ccurate P seudo-labels for S emi-supervised image classification). In BM, the two sub-networks jointly predict the unlabeled images, then select the pseudo-labels with a high-confidence threshold to perform the balancing operation to obtain the initial samples with balanced distribution of each category. In GMM, referring to the common practice of the Learning from Noise Labels task, we use GMM to fit the loss distribution of images with pseudo-labels output by BM, then clean samples and noise samples are divided based on the observation that the loss of correctly labeled images is generally smaller than that of wrongly labeled ones. Through BM and GMM, pseudo-labels with balanced distribution and high accuracy are obtained for the subsequent model training process. Our model has achieved better classification accuracy than most state-of-the-art semi-supervised image classification algorithms on the CIFAR-10/100 and SVHN datasets, and further ablation experiments demonstrate the effectiveness of our BAPS. The source code of BAPS will be available at https://github.com/zhaojianaaa .

Collaborative Learning with Unreliability Adaptation for Semi-Supervised Image Classification

Constructing training goals for unlabeled data is crucial for image classification in the semi-supervised setting. Consistency regularization typically encourages a model to produce consistent predictions with the given training goals, while unreliability adaptation aims to learn the transition probabilities from model predictions to training goals, instead of enforcing their consistency. In this paper, we present a model of Collaborative learning with Unreliability Adaptation (CoUA), in which multiple constituent networks collaboratively learn with each other by adapting their predictions. Toward this end, an additional adaptation module is incorporated into each network to learn a transition probability from its own prediction to that of the paired network. Therefore, the networks can exchange training experience, without being overly sensitive to the unreliability of predictions. To further enhance the collaborative learning, each network is encouraged to produce consistent predictions with the consensus results, while being resistant to the adversarial perturbations against others. Therefore, the networks are able to mutually reinforce each other. We perform extensive experiments on multiple image classification benchmarks to verify the superiority of the co-adaptation based collaborative learning mechanism.

Semisupervised image classification by mutual learning of multiple self‐supervised models

Image classification has been widely adopted by current social media applications. Compared with fully supervised classification, semisupervised classification attracts more attention because it is commonly observed that category labels are only available for a small portion of images while most images on social media platforms do not have labels. To this end, we propose a two-stage semisupervised learning framework. In the first stage, we train two Self-supervised Models (SSMs). One model is initialized by predicting the rotation angles of pretransformed training images and then further trained by the labeled images. The other model is initialized by making consistent predictions for the transformed images in color, shape, and quality from the same sample image, and then further trained by the labeled images. In the second stage, we fuse the two SSMs through deep mutual learning, which enhances each of the two SSMs with the complementary information provided by the other such that the correct prediction could be shared. Experimental results on CIFAR and Caltech-256 data sets demonstrate the effect of the proposed framework.

Retaining Diverse Information in Contrastive Learning Through Multiple Projectors

<i>Contrastive Learning (CL)</i> achieves great success in learning visual representations by comparing two augmented views of the same images. However, this very design removes transformation-dependent visual information from the pre-training, which leads to incomplete representations and is harmful for downstream tasks. It&#x0027;s still an open question to retain such information in the CL pre-training process. In this paper, we propose a <i>Multi-Projector Contrastive Learning (MPCL)</i> to address this issue, which produces multi-view contrastive candidates to retain more comprehensive visual characteristics. In addition, we introduce a contrast regularization to construct multiple projectors as different as possible, thereby facilitating the diversity of preserved information. Finally, to promote a consistent learning process for multi-projector, we design a projector training balance strategy to adjust the learning preference of different projectors. MPCL can be applied to various CL frameworks to effectively protect visual characteristics. Experimental results show that the method performs well on subsequent tasks such as linear and semi-supervised image classification, object detection, and semantic segmentation. Importantly, the visual transformer trained by MPCL improves 2&#x0025; absolute points of linear evaluation beyond the MoCo-v3 on the ImageNet-100 dataset.

Discrepant collaborative training by Sinkhorn divergences

Deep Co-Training algorithms are typically comprised of two distinct and diverse feature extractors that simultaneously attempt to learn task-specific features from the same inputs. Achieving such an objective is, however, not trivial, despite its innocent look. This is because homogeneous networks tend to mimic each other under the collaborative training setup. Keeping this difficulty in mind, we make use of the newly proposed S∈ divergence to encourage diversity between homogeneous networks. The S∈ divergence encapsulates popular measures such as maximum mean discrepancy and the Wasserstein distance under the same umbrella and provides us with a principled, yet simple and straightforward mechanism. Our empirical results in two domains, classification in the presence of noisy labels and semi-supervised image classification, clearly demonstrate the benefits of the proposed framework in learning distinct and diverse features. We show that in these respective settings, we achieve impressive results by a notable margin.

A Large Size Image Classification Method Based on Semi-supervised Learning

Background: Semi-supervised learning in the machine learning community has received widespread attention. Semi-supervised learning can use a small number of tagged samples and a large number of untagged samples for efficient learning. Methods: In 2014, Kim proposed a new semi-supervised learning method: the minimax label propagation (MMLP) method. This method reduces time complexity to O (n), with a smaller computation cost and stronger classification ability than traditional methods. However, classification results are not accurate in large-scale image classifications. Thus, in this paper, we propose a semisupervised image classification method, which is an MMLP-based algorithm. The main idea is threefold: (1) Improving connectivity of image pixels by pixel sampling to reduce the image size, at the same time, reduce the diversity of image characteristics; (2) Using a recall feature to improve the MMLP algorithm; (3) through classification mapping, gaining the classification of the original data from the classification of the data reduction. Results: In the end, our algorithm also gains a minimax path from untagged samples to tagged samples. The experimental results proved that this algorithm is applicable to semi-supervised learning on small-size and that it can also gain better classification results for large-size image at the same time. Conclusion: In our paper, considering the connectivity of the neighboring matrix and the diversity of the characteristics, we used meanshift clustering algorithm, next we will use fuzzy energy clustering on our algorithm. We will study the function of these paths.

Image selection based on autoencoder neural network and application to the semi-supervised image classification

Image selection based on autoencoder neural network and application to the semi-supervised image classification

SubSpace Capsule Network

Convolutional neural networks (CNNs) have become a key asset to most of fields in AI. Despite their successful performance, CNNs suffer from a major drawback. They fail to capture the hierarchy of spatial relation among different parts of an entity. As a remedy to this problem, the idea of capsules was proposed by Hinton. In this paper, we propose the SubSpace Capsule Network (SCN) that exploits the idea of capsule networks to model possible variations in the appearance or implicitly-defined properties of an entity through a group of capsule subspaces instead of simply grouping neurons to create capsules. A capsule is created by projecting an input feature vector from a lower layer onto the capsule subspace using a learnable transformation. This transformation finds the degree of alignment of the input with the properties modeled by the capsule subspace.We show that SCN is a general capsule network that can successfully be applied to both discriminative and generative models without incurring computational overhead compared to CNN during test time. Effectiveness of SCN is evaluated through a comprehensive set of experiments on supervised image classification, semi-supervised image classification and high-resolution image generation tasks using the generative adversarial network (GAN) framework. SCN significantly improves the performance of the baseline models in all 3 tasks.

Adversarial Transformations for Semi-Supervised Learning

We propose a Regularization framework based on Adversarial Transformations (RAT) for semi-supervised learning. RAT is designed to enhance robustness of the output distribution of class prediction for a given data against input perturbation. RAT is an extension of Virtual Adversarial Training (VAT) in such a way that RAT adversraialy transforms data along the underlying data distribution by a rich set of data transformation functions that leave class label invariant, whereas VAT simply produces adversarial additive noises. In addition, we verified that a technique of gradually increasing of perturbation region further improves the robustness. In experiments, we show that RAT significantly improves classification performance on CIFAR-10 and SVHN compared to existing regularization methods under standard semi-supervised image classification settings.

A Novel Broad Learning Model-Based Semi-Supervised Image Classification Method

Broad learning system (BLS) is an effective and efficient incremental learning system without the deep architecture. It has strong feature extraction ability and high computational efficiency. However, it is greatly limited in the applicability of supervised learning. For the collected actual data, more data are unlabeled data and less data are labeled data. To overcome these problems, Fick's law assisted propagation (FLAP) is introduced into the BLS to propose a new semi-supervised classification algorithm, namely FLAP-BLS in this paper. In the FLAP-BLS, the FLAP has the labeled ability from the labeled examples to unlabeled examples, it is used to mark plenty of unlabeled samples by few labeled samples in order to obtain a large number of labeled samples and build the sample data matrix. Then an efficient incremental BLS without deep structure can effectively extract features from large-scale data, it is used to effectively classify the sample matrix. Finally, USPS, MNIST and NORB datasets are selected to validate the effectiveness of the FLAP-BLS. The experiment results show that the FLAP-BLS can effectively classify the few labeled samples and a large of unlabeled samples and obtain classification results with high accuracy, and it has faster classification speed, stronger generalization ability and better stability. The proposed method provides a new method for image classification.

Attention-Based Generative Adversarial Network for Semi-supervised Image Classification

Semi-supervised image classification is one of the areas of interest within the computer vision, which can build better classifiers using a few labeled images and plenty of unlabeled images. Recently, semi-supervised image classification methods based on the generative adversarial network (GAN) get promising results. In this paper, we introduce a self-attention mechanism to propose an attention-based GAN for semi-supervised image classification, which can capture global dependencies and adaptively extract important information. Furthermore, we apply spectral normalization, which can stabilize the training of attention-based GAN. We also adopt manifold regularization as an additional regularization term so that we can make the most of the unlabeled images. We test the proposed method on SVHN and CIFAR-10 datasets. The experimental results show that the proposed method is comparable with the state-of-the-art GAN-based semi-supervised image classification methods.

Unsupervised and Semi-Supervised Image Classification With Weak Semantic Consistency

Supervised methods have been widely used for image classifications. Although great progress has been made, existing supervised methods rely on well-labeled samples for classification. However, we often have large quantities of images with few or no labels. To cope with this problem, in this paper, we propose a novel weak semantic consistency constrained image classification method. We start from an extreme circumstance by viewing each image as one class. We train exemplar classifiers to separate each image from other images. For each image, we use the learned exemplar classifiers to predict the weak semantic correlations with the exemplar classifiers. When no labeled information is available, we cluster images using the weak semantic correlations and assign images within one cluster to the same mid-level class. When partially labeled images are available, we can use them to constrain the clustering process by assigning images of varied semantics to different mid-level classes. We use the newly assigned images for classifier training and new image representations, which can then be used for similar image assignments. The classifier training, image representation, and assignment processes are repeated until convergence. We conduct both unsupervised and semi-supervised image classification experiments on several datasets. The experimental results show the effectiveness of the proposed unsupervised and semi-supervised weak semantic consistency image classification method.

Correntropy-Induced Robust Low-Rank Hypergraph.

Hypergraph learning has been widely exploited in various image processing applications, due to its advantages in modeling the high-order information. Its efficacy highly depends on building an informative hypergraph structure to accurately and robustly formulate the underlying data correlation. However, the existing hypergraph learning methods are sensitive to non- Gaussian noise, which hurts the corresponding performance. In this paper, we present a noise-resistant hypergraph learning model, which provides superior robustness against various non- Gaussian noises. In particular, our model adopts low-rank representation to construct a hypergraph, which captures the globally linear data structure as well as preserving the grouping effect of highly-correlated data. We further introduce a correntropyinduced local metric to measure the reconstruction errors, which is particularly robust to non-Gaussian noises. Finally, the Frobenious-norm based regularization is proposed to combine with the low-rank regularizer, which enables our model to regularize the singular values of the coefficient matrix. By such, the non-zero coefficients are selected to generate a hyperedge set as well as the hyperedge weights. We have evaluated the proposed hypergraph model in the tasks of image clustering and semi-supervised image classification. Quantitatively, our scheme significantly enhances the performance of the state-of-the-art hypergraph models on several benchmark datasets.