Digital Breast Tomosynthesis (DBT) is a widely used medical imaging modality for breast cancer screening and diagnosis, offering higher spatial resolution and greater detail through its 3D-like breast volume imaging capability. However, the increased data volume also introduces pronounced data imbalance challenges, where only a small fraction of the volume contains suspicious tissue. This further exacerbates the data imbalance due to the case-level distribution in real-world data and leads to learning a trivial classification model that only predicts the majority class. To address this, we propose a novel method using view-level contrastive Self-supervised Initialization and Fine-Tuning for identifying abnormal DBT images, namely SIFT-DBT. We further introduce a patch-level multi-instance learning method to preserve spatial resolution. The proposed method achieves 92.69% volume-wise AUC on an evaluation of 970 unique studies.
This paper introduces FUNGI, Features from UNsupervised GradIents, a method to enhance the features of transformer encoders by leveraging self-supervised gradients. Our method is simple: given any pretrained model, we first compute gradients from various self-supervised objectives for each input. These gradients are projected to a lower dimension and then concatenated with the model's output embedding. The resulting features are evaluated on k-nearest neighbor classification over 11 datasets from vision, 5 from natural language processing, and 2 from audio. Across backbones spanning various sizes and pretraining strategies, FUNGI features provide consistent performance improvements over the embeddings. We also show that using FUNGI features can benefit linear classification, clustering and image retrieval, and that they significantly improve the retrieval-based in-context scene understanding abilities of pretrained models, for example improving upon DINO by +17% for semantic segmentation - without any training.
Black-box adversarial attacks can fool image classifiers into misclassifying images without requiring access to model structure and weights. Recent studies have reported attack success rates of over 95% with less than 1,000 queries. The question then arises of whether black-box attacks have become a real threat against IoT devices that rely on cloud APIs to achieve image classification. To shed some light on this, note that prior research has primarily focused on increasing the success rate and reducing the number of queries. However, another crucial factor for black-box attacks against cloud APIs is the time required to perform the attack. This paper applies black-box attacks directly to cloud APIs rather than to local models, thereby avoiding mistakes made in prior research that applied the perturbation before image encoding and pre-processing. Further, we exploit load balancing to enable distributed black-box attacks that can reduce the attack time by a factor of about five for both local search and gradient estimation methods.
Imageability (potential of text to evoke a mental image) and concreteness (perceptibility of text) are two psycholinguistic properties that link visual and semantic spaces. It is little surprise that computational methods that estimate them do so using parallel visual and semantic spaces, such as collections of image-caption pairs or multi-modal models. In this paper, we work on the supposition that text itself in an image-caption dataset offers sufficient signals to accurately estimate these properties. We hypothesize, in particular, that the peakedness of the neighborhood of a word in the semantic embedding space reflects its degree of imageability and concreteness. We then propose an unsupervised, distribution-free measure, which we call Neighborhood Stability Measure (NSM), that quantifies the sharpness of peaks. Extensive experiments show that NSM correlates more strongly with ground-truth ratings than existing unsupervised methods, and is a strong predictor of these properties for classification. Our code and data are available on GitHub (https://github.com/Artificial-Memory-Lab/imageability).
Large, pretrained models are commonly finetuned with imagery that is heavily augmented to mimic different conditions and scales, with the resulting models used for various tasks with imagery from a range of spatial scales. Such models overlook scale-specific information in the data for scale-dependent domains, such as remote sensing. In this paper, we present Scale-MAE, a pretraining method that explicitly learns relationships between data at different, known scales throughout the pretraining process. Scale-MAE pretrains a network by masking an input image at a known input scale, where the area of the Earth covered by the image determines the scale of the ViT positional encoding, not the image resolution. Scale-MAE encodes the masked image with a standard ViT backbone, and then decodes the masked image through a bandpass filter to reconstruct low/high frequency images at lower/higher scales. We find that tasking the network with reconstructing both low/high frequency images leads to robust multiscale representations for remote sensing imagery. Scale-MAE achieves an average of a $2.4 - 5.6\%$ non-parametric kNN classification improvement across eight remote sensing datasets compared to current state-of-the-art and obtains a $0.9$ mIoU to $1.7$ mIoU improvement on the SpaceNet building segmentation transfer task for a range of evaluation scales.
Deep neural networks have been shown to be susceptible to adversarial examples -- small, imperceptible changes constructed to cause mis-classification in otherwise highly accurate image classifiers. As a practical alternative, recent work proposed so-called adversarial patches: clearly visible, but adversarially crafted rectangular patches in images. These patches can easily be printed and applied in the physical world. While defenses against imperceptible adversarial examples have been studied extensively, robustness against adversarial patches is poorly understood. In this work, we first devise a practical approach to obtain adversarial patches while actively optimizing their location within the image. Then, we apply adversarial training on these location-optimized adversarial patches and demonstrate significantly improved robustness on CIFAR10 and GTSRB. Additionally, in contrast to adversarial training on imperceptible adversarial examples, our adversarial patch training does not reduce accuracy.
Large vision-language models (VLMs) are shown to learn rich joint image-text representations enabling high performances in relevant downstream tasks. However, they fail to showcase their quantitative understanding of objects, and they lack good counting-aware representation. This paper conducts a reproducibility study of 'Teaching CLIP to Count to Ten' (Paiss et al., 2023), which presents a method to finetune a CLIP model (Radford et al., 2021) to improve zero-shot counting accuracy in an image while maintaining the performance for zero-shot classification by introducing a counting-contrastive loss term. We improve the model's performance on a smaller subset of their training data with lower computational resources. We verify these claims by reproducing their study with our own code. The implementation can be found at https://github.com/SforAiDl/CountCLIP.
We characterize convolutional neural networks with respect to the relative amount of features per layer. Using a skew normal distribution as a parametrized framework, we investigate the common assumption of monotonously increasing feature-counts with higher layers of architecture designs. Our evaluation on models with VGG-type layers on the MNIST, Fashion-MNIST and CIFAR-10 image classification benchmarks provides evidence that motivates rethinking of our common assumption: architectures that favor larger early layers seem to yield better accuracy.
Synthetic aperture radar (SAR) image change detection is a critical yet challenging task in the field of remote sensing image analysis. The task is non-trivial due to the following challenges: Firstly, intrinsic speckle noise of SAR images inevitably degrades the neural network because of error gradient accumulation. Furthermore, the correlation among various levels or scales of feature maps is difficult to be achieved through summation or concatenation. Toward this end, we proposed a siamese adaptive fusion network for SAR image change detection. To be more specific, two-branch CNN is utilized to extract high-level semantic features of multitemporal SAR images. Besides, an adaptive fusion module is designed to adaptively combine multiscale responses in convolutional layers. Therefore, the complementary information is exploited, and feature learning in change detection is further improved. Moreover, a correlation layer is designed to further explore the correlation between multitemporal images. Thereafter, robust feature representation is utilized for classification through a fully-connected layer with softmax. Experimental results on four real SAR datasets demonstrate that the proposed method exhibits superior performance against several state-of-the-art methods. Our codes are available at https://github.com/summitgao/SAR_CD_SAFNet.
Adversarial training (AT) is a simple yet effective defense against adversarial attacks to image classification systems, which is based on augmenting the training set with attacks that maximize the loss. However, the effectiveness of AT as a defense for video classification has not been thoroughly studied. Our first contribution is to show that generating optimal attacks for video requires carefully tuning the attack parameters, especially the step size. Notably, we show that the optimal step size varies linearly with the attack budget. Our second contribution is to show that using a smaller (sub-optimal) attack budget at training time leads to a more robust performance at test time. Based on these findings, we propose three defenses against attacks with variable attack budgets. The first one, Adaptive AT, is a technique where the attack budget is drawn from a distribution that is adapted as training iterations proceed. The second, Curriculum AT, is a technique where the attack budget is increased as training iterations proceed. The third, Generative AT, further couples AT with a denoising generative adversarial network to boost robust performance. Experiments on the UCF101 dataset demonstrate that the proposed methods improve adversarial robustness against multiple attack types.
Mixup is a data augmentation technique that enhances model generalization by interpolating between data points using a mixing ratio $\lambda$ in the image domain. Recently, the concept of mixup has been adapted to the graph domain through node-centric interpolations. However, these approaches often fail to address the complexity of interconnected relationships, potentially damaging the graph's natural topology and undermining node interactions. Furthermore, current graph mixup methods employ a one-size-fits-all strategy with a randomly sampled $\lambda$ for all mixup pairs, ignoring the diverse needs of different pairs. This paper proposes an Adaptive Graph Mixup (AGMixup) framework for semi-supervised node classification. AGMixup introduces a subgraph-centric approach, which treats each subgraph similarly to how images are handled in Euclidean domains, thus facilitating a more natural integration of mixup into graph-based learning. We also propose an adaptive mechanism to tune the mixing ratio $\lambda$ for diverse mixup pairs, guided by the contextual similarity and uncertainty of the involved subgraphs. Extensive experiments across seven datasets on semi-supervised node classification benchmarks demonstrate AGMixup's superiority over state-of-the-art graph mixup methods. Source codes are available at \url{https://github.com/WeigangLu/AGMixup}.
Automated chest radiographs interpretation requires both accurate disease classification and detailed radiology report generation, presenting a significant challenge in the clinical workflow. Current approaches either focus on classification accuracy at the expense of interpretability or generate detailed but potentially unreliable reports through image captioning techniques. In this study, we present RadAlign, a novel framework that combines the predictive accuracy of vision-language models (VLMs) with the reasoning capabilities of large language models (LLMs). Inspired by the radiologist's workflow, RadAlign first employs a specialized VLM to align visual features with key medical concepts, achieving superior disease classification with an average AUC of 0.885 across multiple diseases. These recognized medical conditions, represented as text-based concepts in the aligned visual-language space, are then used to prompt LLM-based report generation. Enhanced by a retrieval-augmented generation mechanism that grounds outputs in similar historical cases, RadAlign delivers superior report quality with a GREEN score of 0.678, outperforming state-of-the-art methods' 0.634. Our framework maintains strong clinical interpretability while reducing hallucinations, advancing automated medical imaging and report analysis through integrated predictive and generative AI. Code is available at https://github.com/difeigu/RadAlign.
With the increasing attention to deep neural network (DNN) models, attacks are also upcoming for such models. For example, an attacker may carefully construct images in specific ways (also referred to as adversarial examples) aiming to mislead the DNN models to output incorrect classification results. Similarly, many efforts are proposed to detect and mitigate adversarial examples, usually for certain dedicated attacks. In this paper, we propose a novel digital watermark-based method to generate image adversarial examples to fool DNN models. Specifically, partial main features of the watermark image are embedded into the host image almost invisibly, aiming to tamper with and damage the recognition capabilities of the DNN models. We devise an efficient mechanism to select host images and watermark images and utilize the improved discrete wavelet transform (DWT) based Patchwork watermarking algorithm with a set of valid hyperparameters to embed digital watermarks from the watermark image dataset into original images for generating image adversarial examples. The experimental results illustrate that the attack success rate on common DNN models can reach an average of 95.47% on the CIFAR-10 dataset and the highest at 98.71%. Besides, our scheme is able to generate a large number of adversarial examples efficiently, concretely, an average of 1.17 seconds for completing the attacks on each image on the CIFAR-10 dataset. In addition, we design a baseline experiment using the watermark images generated by Gaussian noise as the watermark image dataset that also displays the effectiveness of our scheme. Similarly, we also propose the modified discrete cosine transform (DCT) based Patchwork watermarking algorithm. To ensure repeatability and reproducibility, the source code is available on GitHub.
Evaluating neural network performance is critical to deep neural network design but a costly procedure. Neural predictors provide an efficient solution by treating architectures as samples and learning to estimate their performance on a given task. However, existing predictors are task-dependent, predominantly estimating neural network performance on image classification benchmarks. They are also search-space dependent; each predictor is designed to make predictions for a specific architecture search space with predefined topologies and set of operations. In this paper, we propose a novel All-in-One Predictor (AIO-P), which aims to pretrain neural predictors on architecture examples from multiple, separate computer vision (CV) task domains and multiple architecture spaces, and then transfer to unseen downstream CV tasks or neural architectures. We describe our proposed techniques for general graph representation, efficient predictor pretraining and knowledge infusion techniques, as well as methods to transfer to downstream tasks/spaces. Extensive experimental results show that AIO-P can achieve Mean Absolute Error (MAE) and Spearman's Rank Correlation (SRCC) below 1% and above 0.5, respectively, on a breadth of target downstream CV tasks with or without fine-tuning, outperforming a number of baselines. Moreover, AIO-P can directly transfer to new architectures not seen during training, accurately rank them and serve as an effective performance estimator when paired with an algorithm designed to preserve performance while reducing FLOPs.
Counterfactual explanations (CEs) aim to enhance the interpretability of machine learning models by illustrating how alterations in input features would affect the resulting predictions. Common CE approaches require an additional model and are typically constrained to binary counterfactuals. In contrast, we propose a novel method that operates directly on the latent space of a generative model, specifically a Diffusion Autoencoder (DAE). This approach offers inherent interpretability by enabling the generation of CEs and the continuous visualization of the model's internal representation across decision boundaries. Our method leverages the DAE's ability to encode images into a semantically rich latent space in an unsupervised manner, eliminating the need for labeled data or separate feature extraction models. We show that these latent representations are helpful for medical condition classification and the ordinal regression of severity pathologies, such as vertebral compression fractures (VCF) and diabetic retinopathy (DR). Beyond binary CEs, our method supports the visualization of ordinal CEs using a linear model, providing deeper insights into the model's decision-making process and enhancing interpretability. Experiments across various medical imaging datasets demonstrate the method's advantages in interpretability and versatility. The linear manifold of the DAE's latent space allows for meaningful interpolation and manipulation, making it a powerful tool for exploring medical image properties. Our code is available at https://doi.org/10.5281/zenodo.13859266.
Over the long history of machine learning, which dates back several decades, recurrent neural networks (RNNs) have been used mainly for sequential data and time series and generally with 1D information. Even in some rare studies on 2D images, these networks are used merely to learn and generate data sequentially rather than for image recognition tasks. In this study, we propose integrating an RNN as an additional layer when designing image recognition models. We also develop end-to-end multimodel ensembles that produce expert predictions using several models. In addition, we extend the training strategy so that our model performs comparably to leading models and can even match the state-of-the-art models on several challenging datasets (e.g., SVHN (0.99), Cifar-100 (0.9027) and Cifar-10 (0.9852)). Moreover, our model sets a new record on the Surrey dataset (0.949). The source code of the methods provided in this article is available at https://github.com/leonlha/e2e-3m and http://nguyenhuuphong.me.
Outcomes with a natural order commonly occur in prediction tasks and often the available input data are a mixture of complex data like images and tabular predictors. Deep Learning (DL) models are state-of-the-art for image classification tasks but frequently treat ordinal outcomes as unordered and lack interpretability. In contrast, classical ordinal regression models consider the outcome's order and yield interpretable predictor effects but are limited to tabular data. We present ordinal neural network transformation models (ONTRAMs), which unite DL with classical ordinal regression approaches. ONTRAMs are a special case of transformation models and trade off flexibility and interpretability by additively decomposing the transformation function into terms for image and tabular data using jointly trained neural networks. The performance of the most flexible ONTRAM is by definition equivalent to a standard multi-class DL model trained with cross-entropy while being faster in training when facing ordinal outcomes. Lastly, we discuss how to interpret model components for both tabular and image data on two publicly available datasets.
State-of-the-art object detection and segmentation methods for microscopy images rely on supervised machine learning, which requires laborious manual annotation of training data. Here we present a self-supervised method based on time arrow prediction pre-training that learns dense image representations from raw, unlabeled live-cell microscopy videos. Our method builds upon the task of predicting the correct order of time-flipped image regions via a single-image feature extractor followed by a time arrow prediction head that operates on the fused features. We show that the resulting dense representations capture inherently time-asymmetric biological processes such as cell divisions on a pixel-level. We furthermore demonstrate the utility of these representations on several live-cell microscopy datasets for detection and segmentation of dividing cells, as well as for cell state classification. Our method outperforms supervised methods, particularly when only limited ground truth annotations are available as is commonly the case in practice. We provide code at https://github.com/weigertlab/tarrow.
Unsupervised learning has always been appealing to machine learning researchers and practitioners, allowing them to avoid an expensive and complicated process of labeling the data. However, unsupervised learning of complex data is challenging, and even the best approaches show much weaker performance than their supervised counterparts. Self-supervised deep learning has become a strong instrument for representation learning in computer vision. However, those methods have not been evaluated in a fully unsupervised setting. In this paper, we propose a simple scheme for unsupervised classification based on self-supervised representations. We evaluate the proposed approach with several recent self-supervised methods showing that it achieves competitive results for ImageNet classification (39% accuracy on ImageNet with 1000 clusters and 46% with overclustering). We suggest adding the unsupervised evaluation to a set of standard benchmarks for self-supervised learning. The code is available at https://github.com/Randl/kmeans_selfsuper
Grounded situation recognition is the task of predicting the main activity, entities playing certain roles within the activity, and bounding-box groundings of the entities in the given image. To effectively deal with this challenging task, we introduce a novel approach where the two processes for activity classification and entity estimation are interactive and complementary. To implement this idea, we propose Collaborative Glance-Gaze TransFormer (CoFormer) that consists of two modules: Glance transformer for activity classification and Gaze transformer for entity estimation. Glance transformer predicts the main activity with the help of Gaze transformer that analyzes entities and their relations, while Gaze transformer estimates the grounded entities by focusing only on the entities relevant to the activity predicted by Glance transformer. Our CoFormer achieves the state of the art in all evaluation metrics on the SWiG dataset. Training code and model weights are available at https://github.com/jhcho99/CoFormer.