Data augmentation is one of the most prevalent tools in deep learning, underpinning many recent advances, including those from classification, generative models, and representation learning. The standard approach to data augmentation combines simple transformations like rotations and flips to generate new images from existing ones. However, these new images lack diversity along key semantic axes present in the data. Current augmentations cannot alter the high-level semantic attributes, such as animal species present in a scene, to enhance the diversity of data. We address the lack of diversity in data augmentation with image-to-image transformations parameterized by pre-trained text-to-image diffusion models. Our method edits images to change their semantics using an off-the-shelf diffusion model, and generalizes to novel visual concepts from a few labelled examples. We evaluate our approach on few-shot image classification tasks, and on a real-world weed recognition task, and observe an improvement in accuracy in tested domains.
The success of Vision Transformer (ViT) in various computer vision tasks has promoted the ever-increasing prevalence of this convolution-free network. The fact that ViT works on image patches makes it potentially relevant to the problem of jigsaw puzzle solving, which is a classical self-supervised task aiming at reordering shuffled sequential image patches back to their natural form. Despite its simplicity, solving jigsaw puzzle has been demonstrated to be helpful for diverse tasks using Convolutional Neural Networks (CNNs), such as self-supervised feature representation learning, domain generalization, and fine-grained classification. In this paper, we explore solving jigsaw puzzle as a self-supervised auxiliary loss in ViT for image classification, named Jigsaw-ViT. We show two modifications that can make Jigsaw-ViT superior to standard ViT: discarding positional embeddings and masking patches randomly. Yet simple, we find that Jigsaw-ViT is able to improve both in generalization and robustness over the standard ViT, which is usually rather a trade-off. Experimentally, we show that adding the jigsaw puzzle branch provides better generalization than ViT on large-scale image classification on ImageNet. Moreover, the auxiliary task also improves robustness to noisy labels on Animal-10N, Food-101N, and Clothing1M as well as adversarial examples. Our implementation is available at https://yingyichen-cyy.github.io/Jigsaw-ViT/.
Hyperspectral image (HSI) classification is widely used for the analysis of remotely sensed images. Hyperspectral imagery includes varying bands of images. Convolutional Neural Network (CNN) is one of the most frequently used deep learning based methods for visual data processing. The use of CNN for HSI classification is also visible in recent works. These approaches are mostly based on 2D CNN. Whereas, the HSI classification performance is highly dependent on both spatial and spectral information. Very few methods have utilized the 3D CNN because of increased computational complexity. This letter proposes a Hybrid Spectral Convolutional Neural Network (HybridSN) for HSI classification. Basically, the HybridSN is a spectral-spatial 3D-CNN followed by spatial 2D-CNN. The 3D-CNN facilitates the joint spatial-spectral feature representation from a stack of spectral bands. The 2D-CNN on top of the 3D-CNN further learns more abstract level spatial representation. Moreover, the use of hybrid CNNs reduces the complexity of the model compared to 3D-CNN alone. To test the performance of this hybrid approach, very rigorous HSI classification experiments are performed over Indian Pines, Pavia University and Salinas Scene remote sensing datasets. The results are compared with the state-of-the-art hand-crafted as well as end-to-end deep learning based methods. A very satisfactory performance is obtained using the proposed HybridSN for HSI classification. The source code can be found at \url{https://github.com/gokriznastic/HybridSN}.
In recent years, hyperspectral image (HSI) classification based on generative adversarial networks (GAN) has achieved great progress. GAN-based classification methods can mitigate the limited training sample dilemma to some extent. However, several studies have pointed out that existing GAN-based HSI classification methods are heavily affected by the imbalanced training data problem. The discriminator in GAN always contradicts itself and tries to associate fake labels to the minority-class samples, and thus impair the classification performance. Another critical issue is the mode collapse in GAN-based methods. The generator is only capable of producing samples within a narrow scope of the data space, which severely hinders the advancement of GAN-based HSI classification methods. In this paper, we proposed an Adaptive DropBlock-enhanced Generative Adversarial Networks (ADGAN) for HSI classification. First, to solve the imbalanced training data problem, we adjust the discriminator to be a single classifier, and it will not contradict itself. Second, an adaptive DropBlock (AdapDrop) is proposed as a regularization method employed in the generator and discriminator to alleviate the mode collapse issue. The AdapDrop generated drop masks with adaptive shapes instead of a fixed size region, and it alleviates the limitations of DropBlock in dealing with ground objects with various shapes. Experimental results on three HSI datasets demonstrated that the proposed ADGAN achieved superior performance over state-of-the-art GAN-based methods. Our codes are available at https://github.com/summitgao/HC_ADGAN
Automated benthic image annotation is crucial to efficiently monitor and protect coral reefs against climate change. Current machine learning approaches fail to capture the hierarchical nature of benthic organisms covering reef substrata, i.e., coral taxonomic levels and health condition. To address this limitation, we propose to annotate benthic images using hierarchical classification. Experiments on a custom dataset from a Northeast Brazilian coral reef show that our approach outperforms flat classifiers, improving both F1 and hierarchical F1 scores by approximately 2\% across varying amounts of training data. In addition, this hierarchical method aligns more closely with ecological objectives.
Ultrasound video classification enables automated diagnosis and has emerged as an important research area. However, publicly available ultrasound video datasets remain scarce, hindering progress in developing effective video classification models. We propose addressing this shortage by synthesizing plausible ultrasound videos from readily available, abundant ultrasound images. To this end, we introduce a latent dynamic diffusion model (LDDM) to efficiently translate static images to dynamic sequences with realistic video characteristics. We demonstrate strong quantitative results and visually appealing synthesized videos on the BUSV benchmark. Notably, training video classification models on combinations of real and LDDM-synthesized videos substantially improves performance over using real data alone, indicating our method successfully emulates dynamics critical for discrimination. Our image-to-video approach provides an effective data augmentation solution to advance ultrasound video analysis. Code is available at https://github.com/MedAITech/U_I2V.
Fine-tuned vision-language models (VLMs) often capture spurious correlations between image features and textual attributes, resulting in degraded zero-shot performance at test time. Existing approaches for addressing spurious correlations (i) primarily operate at the global image-level rather than intervening directly on fine-grained image features and (ii) are predominantly designed for unimodal settings. In this work, we present RaVL, which takes a fine-grained perspective on VLM robustness by discovering and mitigating spurious correlations using local image features rather than operating at the global image level. Given a fine-tuned VLM, RaVL first discovers spurious correlations by leveraging a region-level clustering approach to identify precise image features contributing to zero-shot classification errors. Then, RaVL mitigates the identified spurious correlation with a novel region-aware loss function that enables the VLM to focus on relevant regions and ignore spurious relationships during fine-tuning. We evaluate RaVL on 654 VLMs with various model architectures, data domains, and learned spurious correlations. Our results show that RaVL accurately discovers (191% improvement over the closest baseline) and mitigates (8.2% improvement on worst-group image classification accuracy) spurious correlations. Qualitative evaluations on general-domain and medical-domain VLMs confirm our findings.
We address the challenging problem of whole slide image (WSI) classification. WSIs have very high resolutions and usually lack localized annotations. WSI classification can be cast as a multiple instance learning (MIL) problem when only slide-level labels are available. We propose a MIL-based method for WSI classification and tumor detection that does not require localized annotations. Our method has three major components. First, we introduce a novel MIL aggregator that models the relations of the instances in a dual-stream architecture with trainable distance measurement. Second, since WSIs can produce large or unbalanced bags that hinder the training of MIL models, we propose to use self-supervised contrastive learning to extract good representations for MIL and alleviate the issue of prohibitive memory cost for large bags. Third, we adopt a pyramidal fusion mechanism for multiscale WSI features, and further improve the accuracy of classification and localization. Our model is evaluated on two representative WSI datasets. The classification accuracy of our model compares favorably to fully-supervised methods, with less than 2% accuracy gap across datasets. Our results also outperform all previous MIL-based methods. Additional benchmark results on standard MIL datasets further demonstrate the superior performance of our MIL aggregator on general MIL problems. GitHub repository: https://github.com/binli123/dsmil-wsi
Malignant melanoma (MM) is one of the deadliest types of skin cancer. Analysing dermatoscopic images plays an important role in the early detection of MM and other pigmented skin lesions. Among different computer-based methods, deep learning-based approaches and in particular convolutional neural networks have shown excellent classification and segmentation performances for dermatoscopic skin lesion images. These models can be trained end-to-end without requiring any hand-crafted features. However, the effect of using lesion segmentation information on classification performance has remained an open question. In this study, we explicitly investigated the impact of using skin lesion segmentation masks on the performance of dermatoscopic image classification. To do this, first, we developed a baseline classifier as the reference model without using any segmentation masks. Then, we used either manually or automatically created segmentation masks in both training and test phases in different scenarios and investigated the classification performances. Evaluated on the ISIC 2017 challenge dataset which contained two binary classification tasks (i.e. MM vs. all and seborrheic keratosis (SK) vs. all) and based on the derived area under the receiver operating characteristic curve scores, we observed four main outcomes. Our results show that 1) using segmentation masks did not significantly improve the MM classification performance in any scenario, 2) in one of the scenarios (using segmentation masks for dilated cropping), SK classification performance was significantly improved, 3) removing all background information by the segmentation masks significantly degraded the overall classification performance, and 4) in case of using the appropriate scenario (using segmentation for dilated cropping), there is no significant difference of using manually or automatically created segmentation masks.
General-purpose foundation models have led to recent breakthroughs in artificial intelligence. In remote sensing, self-supervised learning (SSL) and Masked Image Modeling (MIM) have been adopted to build foundation models. However, these models primarily learn low-level features and require annotated data for fine-tuning. Moreover, they are inapplicable for retrieval and zero-shot applications due to the lack of language understanding. To address these limitations, we propose RemoteCLIP, the first vision-language foundation model for remote sensing that aims to learn robust visual features with rich semantics and aligned text embeddings for seamless downstream application. To address the scarcity of pre-training data, we leverage data scaling which converts heterogeneous annotations into a unified image-caption data format based on Box-to-Caption (B2C) and Mask-to-Box (M2B) conversion. By further incorporating UAV imagery, we produce a 12 $\times$ larger pretraining dataset than the combination of all available datasets. RemoteCLIP can be applied to a variety of downstream tasks, including zero-shot image classification, linear probing, $\textit{k}$-NN classification, few-shot classification, image-text retrieval, and object counting in remote sensing images. Evaluation on 16 datasets, including a newly introduced RemoteCount benchmark to test the object counting ability, shows that RemoteCLIP consistently outperforms baseline foundation models across different model scales. Impressively, RemoteCLIP beats the state-of-the-art method by 9.14% mean recall on the RSITMD dataset and 8.92% on the RSICD dataset. For zero-shot classification, our RemoteCLIP outperforms the CLIP baseline by up to 6.39% average accuracy on 12 downstream datasets. Project website: https://github.com/ChenDelong1999/RemoteCLIP
Failure detection in automated image classification is a critical safeguard for clinical deployment. Detected failure cases can be referred to human assessment, ensuring patient safety in computer-aided clinical decision making. Despite its paramount importance, there is insufficient evidence about the ability of state-of-the-art confidence scoring methods to detect test-time failures of classification models in the context of medical imaging. This paper provides a reality check, establishing the performance of in-domain misclassification detection methods, benchmarking 9 widely used confidence scores on 6 medical imaging datasets with different imaging modalities, in multiclass and binary classification settings. Our experiments show that the problem of failure detection is far from being solved. We found that none of the benchmarked advanced methods proposed in the computer vision and machine learning literature can consistently outperform a simple softmax baseline, demonstrating that improved out-of-distribution detection or model calibration do not necessarily translate to improved in-domain misclassification detection. Our developed testbed facilitates future work in this important area
Modern image classification systems are often built on deep neural networks, which suffer from adversarial examples--images with deliberately crafted, imperceptible noise to mislead the network's classification. To defend against adversarial examples, a plausible idea is to obfuscate the network's gradient with respect to the input image. This general idea has inspired a long line of defense methods. Yet, almost all of them have proven vulnerable. We revisit this seemingly flawed idea from a radically different perspective. We embrace the omnipresence of adversarial examples and the numerical procedure of crafting them, and turn this harmful attacking process into a useful defense mechanism. Our defense method is conceptually simple: before feeding an input image for classification, transform it by finding an adversarial example on a pre-trained external model. We evaluate our method against a wide range of possible attacks. On both CIFAR-10 and Tiny ImageNet datasets, our method is significantly more robust than state-of-the-art methods. Particularly, in comparison to adversarial training, our method offers lower training cost as well as stronger robustness.
We present a study of the potential for Convolutional Neural Networks (CNNs) to enable separation of astrophysical transients from image artifacts, a task known as "real-bogus" classification without requiring a template subtracted (or difference) image which requires a computationally expensive process to generate, involving image matching on small spatial scales in large volumes of data. Using data from the Dark Energy Survey, we explore the use of CNNs to (1) automate the "real-bogus" classification, (2) reduce the computational costs of transient discovery. We compare the efficiency of two CNNs with similar architectures, one that uses "image triplets" (templates, search, and difference image) and one that takes as input the template and search only. We measure the decrease in efficiency associated with the loss of information in input finding that the testing accuracy is reduced from 96% to 91.1%. We further investigate how the latter model learns the required information from the template and search by exploring the saliency maps. Our work (1) confirms that CNNs are excellent models for "real-bogus" classification that rely exclusively on the imaging data and require no feature engineering task; (2) demonstrates that high-accuracy (> 90%) models can be built without the need to construct difference images, but some accuracy is lost. Since once trained, neural networks can generate predictions at minimal computational costs, we argue that future implementations of this methodology could dramatically reduce the computational costs in the detection of transients in synoptic surveys like Rubin Observatory's Legacy Survey of Space and Time by bypassing the Difference Image Analysis entirely.
Weakly supervised learning has been rapidly advanced in biomedical image analysis to achieve pixel-wise labels (segmentation) from image-wise annotations (classification), as biomedical images naturally contain image-wise labels in many scenarios. The current weakly supervised learning algorithms from the computer vision community are largely designed for focal objects (e.g., dogs and cats). However, such algorithms are not optimized for diffuse patterns in biomedical imaging (e.g., stains and fluorescence in microscopy imaging). In this paper, we propose a novel class-aware codebook learning (CaCL) algorithm to perform weakly supervised learning for diffuse image patterns. Specifically, the CaCL algorithm is deployed to segment protein expressed brush border regions from histological images of human duodenum. Our contribution is three-fold: (1) we approach the weakly supervised segmentation from a novel codebook learning perspective; (2) the CaCL algorithm segments diffuse image patterns rather than focal objects; and (3) the proposed algorithm is implemented in a multi-task framework based on Vector Quantised-Variational AutoEncoder (VQ-VAE) via joint image reconstruction, classification, feature embedding, and segmentation. The experimental results show that our method achieved superior performance compared with baseline weakly supervised algorithms. The code is available at https://github.com/ddrrnn123/CaCL.
With advances in digital technology, the classification of medical images has become a crucial step for image-based clinical decision support systems. Automatic medical image classification represents a pivotal domain where the use of AI holds the potential to create a significant social impact. However, several challenges act as obstacles to the development of practical and effective solutions. One of these challenges is the prevalent class imbalance problem in most medical imaging datasets. As a result, existing AI techniques, particularly deep-learning-based methodologies, often underperform in such scenarios. In this study, we propose a novel framework called Large Margin aware Focal (LMF) loss to mitigate the class imbalance problem in medical imaging. The LMF loss represents a linear combination of two loss functions optimized by two hyperparameters. This framework harnesses the distinct characteristics of both loss functions by enforcing wider margins for minority classes while simultaneously emphasizing challenging samples found in the datasets. We perform rigorous experiments on three neural network architectures and with four medical imaging datasets. We provide empirical evidence that our proposed framework consistently outperforms other baseline methods, showing an improvement of 2%-9% in macro-f1 scores. Through class-wise analysis of f1 scores, we also demonstrate how the proposed framework can significantly improve performance for minority classes. The results of our experiments show that our proposed framework can perform consistently well across different architectures and datasets. Overall, our study demonstrates a simple and effective approach to addressing the class imbalance problem in medical imaging datasets. We hope our work will inspire new research toward a more generalized approach to medical image classification.
We address the problem of learning self-supervised representations from unlabeled image collections. Unlike existing approaches that attempt to learn useful features by maximizing similarity between augmented versions of each input image or by speculatively picking negative samples, we instead also make use of the natural variation that occurs in image collections that are captured using static monitoring cameras. To achieve this, we exploit readily available context data that encodes information such as the spatial and temporal relationships between the input images. We are able to learn representations that are surprisingly effective for downstream supervised classification, by first identifying high probability positive pairs at training time, i.e. those images that are likely to depict the same visual concept. For the critical task of global biodiversity monitoring, this results in image features that can be adapted to challenging visual species classification tasks with limited human supervision. We present results on four different camera trap image collections, across three different families of self-supervised learning methods, and show that careful image selection at training time results in superior performance compared to existing baselines such as conventional self-supervised training and transfer learning.
Since the development of self-supervised visual representation learning from contrastive learning to masked image modeling (MIM), there is no significant difference in essence, that is, how to design proper pretext tasks for vision dictionary look-up. MIM recently dominates this line of research with state-of-the-art performance on vision Transformers (ViTs), where the core is to enhance the patch-level visual context capturing of the network via denoising auto-encoding mechanism. Rather than tailoring image tokenizers with extra training stages as in previous works, we unleash the great potential of contrastive learning on denoising auto-encoding and introduce a pure MIM method, ConMIM, to produce simple intra-image inter-patch contrastive constraints as the sole learning objectives for masked patch prediction. We further strengthen the denoising mechanism with asymmetric designs, including image perturbations and model progress rates, to improve the network pre-training. ConMIM-pretrained models with various scales achieve competitive results on downstream image classification, semantic segmentation, object detection, and instance segmentation tasks, e.g., on ImageNet-1K classification, we achieve 83.9% top-1 accuracy with ViT-Small and 85.3% with ViT-Base without extra data for pre-training.
Deep learning for histopathology has been successfully used for disease classification, image segmentation and more. However, combining image and text modalities using current state-of-the-art methods has been a challenge due to the high resolution of histopathology images. Automatic report generation for histopathology images is one such challenge. In this work, we show that using an existing pre-trained Vision Transformer in a two-step process of first using it to encode 4096x4096 sized patches of the Whole Slide Image (WSI) and then using it as the encoder and an LSTM decoder for report generation, we can build a fairly performant and portable report generation mechanism that takes into account the whole of the high resolution image, instead of just the patches. We are also able to use representations from an existing powerful pre-trained hierarchical vision transformer and show its usefulness in not just zero shot classification but also for report generation.
In this paper we investigate the use of Fourier Neural Operators (FNOs) for image classification in comparison to standard Convolutional Neural Networks (CNNs). Neural operators are a discretization-invariant generalization of neural networks to approximate operators between infinite dimensional function spaces. FNOs - which are neural operators with a specific parametrization - have been applied successfully in the context of parametric PDEs. We derive the FNO architecture as an example for continuous and Fr\'echet-differentiable neural operators on Lebesgue spaces. We further show how CNNs can be converted into FNOs and vice versa and propose an interpolation-equivariant adaptation of the architecture.
Federated learning enables building a shared model from multicentre data while storing the training data locally for privacy. In this paper, we present an evaluation (called CXR-FL) of deep learning-based models for chest X-ray image analysis using the federated learning method. We examine the impact of federated learning parameters on the performance of central models. Additionally, we show that classification models perform worse if trained on a region of interest reduced to segmentation of the lung compared to the full image. However, focusing training of the classification model on the lung area may result in improved pathology interpretability during inference. We also find that federated learning helps maintain model generalizability. The pre-trained weights and code are publicly available at (https://github.com/SanoScience/CXR-FL).