arXiv subjects

Nicola Strisciuglio

Publications and source records attributed to Nicola Strisciuglio.

At least 19 recordsLinked to original sources

arXiv2026

Inhibited Self-Attention: Sharpening Focus in Vision Transformers

Vision Transformers (ViTs) have demonstrated remarkable performance in computer vision tasks. However, their self-attention mechanism often diffuses focus across background regions, relying on spurious correlations rather than object-relevant cues. Inspired by inhibitory mechanisms observed in biological vision systems, we propose the Inhibited Self-Attention (ISA), a novel self-attention that integrates inhibitory signals to enhance feature selectivity and suppress spurious responses. In contrast to conventional self-attention, which relies solely on positive attention values due to softmax normalization, our approach retains and utilizes negative attention scores to suppress irrelevant features and sharpen focus on objects of interest. Experiments across multiple datasets, including ImageNet-1k and COCO, and several robustness benchmarks demonstrate that ISA enhances object-centric selectivity, reduces shortcut reliance, and improves out-of-distribution generalization. Our analysis of relevance maps confirms that ViTs with ISA exhibit sharper, more localized focus on object-relevant regions while reducing distractions from non-relevant (background) features, enabling more reliable models. We release our code at https://github.com/prdvanderwal/inhibited-self-attention

cs.CV↗

arXiv2025

Do ImageNet-trained models learn shortcuts? The impact of frequency shortcuts on generalization

Frequency shortcuts refer to specific frequency patterns that models heavily rely on for correct classification. Previous studies have shown that models trained on small image datasets often exploit such shortcuts, potentially impairing their generalization performance. However, existing methods for identifying frequency shortcuts require expensive computations and become impractical for analyzing models trained on large datasets. In this work, we propose the first approach to more efficiently analyze frequency shortcuts at a large scale. We show that both CNN and transformer models learn frequency shortcuts on ImageNet. We also expose that frequency shortcut solutions can yield good performance on out-of-distribution (OOD) test sets which largely retain texture information. However, these shortcuts, mostly aligned with texture patterns, hinder model generalization on rendition-based OOD test sets. These observations suggest that current OOD evaluations often overlook the impact of frequency shortcuts on model generalization. Future benchmarks could thus benefit from explicitly assessing and accounting for these shortcuts to build models that generalize across a broader range of OOD scenarios.

cs.CV↗

arXiv2025

Not Only Text: Exploring Compositionality of Visual Representations in Vision-Language Models

Vision-Language Models (VLMs) learn a shared feature space for text and images, enabling the comparison of inputs of different modalities. While prior works demonstrated that VLMs organize natural language representations into regular structures encoding composite meanings, it remains unclear if compositional patterns also emerge in the visual embedding space. In this work, we investigate compositionality in the image domain, where the analysis of compositional properties is challenged by noise and sparsity of visual data. We address these problems and propose a framework, called Geodesically Decomposable Embeddings (GDE), that approximates image representations with geometry-aware compositional structures in the latent space. We demonstrate that visual embeddings of pre-trained VLMs exhibit a compositional arrangement, and evaluate the effectiveness of this property in the tasks of compositional classification and group robustness. GDE achieves stronger performance in compositional classification compared to its counterpart method that assumes linear geometry of the latent space. Notably, it is particularly effective for group robustness, where we achieve higher results than task-specific solutions. Our results indicate that VLMs can automatically develop a human-like form of compositional reasoning in the visual domain, making their underlying processes more interpretable. Code is available at https://github.com/BerasiDavide/vlm_image_compositionality.

cs.CV↗

arXiv2025

Dynamic Sparse Training versus Dense Training: The Unexpected Winner in Image Corruption Robustness

It is generally perceived that Dynamic Sparse Training opens the door to a new era of scalability and efficiency for artificial neural networks at, perhaps, some costs in accuracy performance for the classification task. At the same time, Dense Training is widely accepted as being the "de facto" approach to train artificial neural networks if one would like to maximize their robustness against image corruption. In this paper, we question this general practice. Consequently, we claim that, contrary to what is commonly thought, the Dynamic Sparse Training methods can consistently outperform Dense Training in terms of robustness accuracy, particularly if the efficiency aspect is not considered as a main objective (i.e., sparsity levels between 10% and up to 50%), without adding (or even reducing) resource cost. We validate our claim on two types of data, images and videos, using several traditional and modern deep learning architectures for computer vision and three widely studied Dynamic Sparse Training algorithms. Our findings reveal a new yet-unknown benefit of Dynamic Sparse Training and open new possibilities in improving deep learning robustness beyond the current state of the art.

cs.CV↗

arXiv2024

Case-level Breast Cancer Prediction for Real Hospital Settings

Breast cancer prediction models for mammography assume that annotations are available for individual images or regions of interest (ROIs), and that there is a fixed number of images per patient. These assumptions do not hold in real hospital settings, where clinicians provide only a final diagnosis for the entire mammography exam (case). Since data in real hospital settings scales with continuous patient intake, while manual annotation efforts do not, we develop a framework for case-level breast cancer prediction that does not require any manual annotation and can be trained with case labels readily available at the hospital. Specifically, we propose a two-level multi-instance learning (MIL) approach at patch and image level for case-level breast cancer prediction and evaluate it on two public and one private dataset. We propose a novel domain-specific MIL pooling observing that breast cancer may or may not occur in both sides, while images of both breasts are taken as a precaution during mammography. We propose a dynamic training procedure for training our MIL framework on a variable number of images per case. We show that our two-level MIL model can be applied in real hospital settings where only case labels, and a variable number of images per case are available, without any loss in performance compared to models trained on image labels. Only trained with weak (case-level) labels, it has the capability to point out in which breast side, mammography view and view region the abnormality lies.

cs.CV↗

arXiv2024

A Survey on the Robustness of Computer Vision Models against Common Corruptions

The performance of computer vision models are susceptible to unexpected changes in input images caused by sensor errors or extreme imaging environments, known as common corruptions (e.g. noise, blur, illumination changes). These corruptions can significantly hinder the reliability of these models when deployed in real-world scenarios, yet they are often overlooked when testing model generalization and robustness. In this survey, we present a comprehensive overview of methods that improve the robustness of computer vision models against common corruptions. We categorize methods into three groups based on the model components and training methods they target: data augmentation, learning strategies, and network components. We release a unified benchmark framework (available at \url{https://github.com/nis-research/CorruptionBenchCV}) to compare robustness performance across several datasets, and we address the inconsistencies of evaluation practices in the literature. Our experimental analysis highlights the base corruption robustness of popular vision backbones, revealing that corruption robustness does not necessarily scale with model size and data size. Large models gain negligible robustness improvements, considering the increased computational requirements. To achieve generalizable and robust computer vision models, we foresee the need of developing new learning strategies that efficiently exploit limited data and mitigate unreliable learning behaviors.

cs.CV↗

arXiv2024

PushPull-Net: Inhibition-driven ResNet robust to image corruptions

We introduce a novel computational unit, termed PushPull-Conv, in the first layer of a ResNet architecture, inspired by the anti-phase inhibition phenomenon observed in the primary visual cortex. This unit redefines the traditional convolutional layer by implementing a pair of complementary filters: a trainable push kernel and its counterpart, the pull kernel. The push kernel (analogous to traditional convolution) learns to respond to specific stimuli, while the pull kernel reacts to the same stimuli but of opposite contrast. This configuration enhances stimulus selectivity and effectively inhibits response in regions lacking preferred stimuli. This effect is attributed to the push and pull kernels, which produce responses of comparable magnitude in such regions, thereby neutralizing each other. The incorporation of the PushPull-Conv into ResNets significantly increases their robustness to image corruption. Our experiments with benchmark corruption datasets show that the PushPull-Conv can be combined with other data augmentation techniques to further improve model robustness. We set a new robustness benchmark on ResNet50 achieving an $mCE$ of 49.95$\%$ on ImageNet-C when combining PRIME augmentation with PushPull inhibition.

cs.CV↗

arXiv2024

Indoor scene recognition from images under visual corruptions

The classification of indoor scenes is a critical component in various applications, such as intelligent robotics for assistive living. While deep learning has significantly advanced this field, models often suffer from reduced performance due to image corruption. This paper presents an innovative approach to indoor scene recognition that leverages multimodal data fusion, integrating caption-based semantic features with visual data to enhance both accuracy and robustness against corruption. We examine two multimodal networks that synergize visual features from CNN models with semantic captions via a Graph Convolutional Network (GCN). Our study shows that this fusion markedly improves model performance, with notable gains in Top-1 accuracy when evaluated against a corrupted subset of the Places365 dataset. Moreover, while standalone visual models displayed high accuracy on uncorrupted images, their performance deteriorated significantly with increased corruption severity. Conversely, the multimodal models demonstrated improved accuracy in clean conditions and substantial robustness to a range of image corruptions. These results highlight the efficacy of incorporating high-level contextual information through captions, suggesting a promising direction for enhancing the resilience of classification systems.

cs.CV↗

arXiv2024

RDA-INR: Riemannian Diffeomorphic Autoencoding via Implicit Neural Representations

Diffeomorphic registration frameworks such as Large Deformation Diffeomorphic Metric Mapping (LDDMM) are used in computer graphics and the medical domain for atlas building, statistical latent modeling, and pairwise and groupwise registration. In recent years, researchers have developed neural network-based approaches regarding diffeomorphic registration to improve the accuracy and computational efficiency of traditional methods. In this work, we focus on a limitation of neural network-based atlas building and statistical latent modeling methods, namely that they either are (i) resolution dependent or (ii) disregard any data- or problem-specific geometry needed for proper mean-variance analysis. In particular, we overcome this limitation by designing a novel encoder based on resolution-independent implicit neural representations. The encoder achieves resolution invariance for LDDMM-based statistical latent modeling. Additionally, the encoder adds LDDMM Riemannian geometry to resolution-independent deep learning models for statistical latent modeling. We investigate how the Riemannian geometry improves latent modeling and is required for a proper mean-variance analysis. To highlight the benefit of resolution independence for LDDMM-based data variability modeling, we show that our approach outperforms current neural network-based LDDMM latent code models. Our work paves the way for more research into how Riemannian geometry, shape respectively image analysis, and deep learning can be combined.

eess.IV↗

arXiv2024

Fourier-basis Functions to Bridge Augmentation Gap: Rethinking Frequency Augmentation in Image Classification

Computer vision models normally witness degraded performance when deployed in real-world scenarios, due to unexpected changes in inputs that were not accounted for during training. Data augmentation is commonly used to address this issue, as it aims to increase data variety and reduce the distribution gap between training and test data. However, common visual augmentations might not guarantee extensive robustness of computer vision models. In this paper, we propose Auxiliary Fourier-basis Augmentation (AFA), a complementary technique targeting augmentation in the frequency domain and filling the augmentation gap left by visual augmentations. We demonstrate the utility of augmentation via Fourier-basis additive noise in a straightforward and efficient adversarial setting. Our results show that AFA benefits the robustness of models against common corruptions, OOD generalization, and consistency of performance of models against increasing perturbations, with negligible deficit to the standard performance of models. It can be seamlessly integrated with other augmentation techniques to further boost performance. Code and models can be found at: https://github.com/nis-research/afa-augment

cs.CV↗

arXiv2024

Quantifying white matter hyperintensity and brain volumes in heterogeneous clinical and low-field portable MRI

Brain atrophy and white matter hyperintensity (WMH) are critical neuroimaging features for ascertaining brain injury in cerebrovascular disease and multiple sclerosis. Automated segmentation and quantification is desirable but existing methods require high-resolution MRI with good signal-to-noise ratio (SNR). This precludes application to clinical and low-field portable MRI (pMRI) scans, thus hampering large-scale tracking of atrophy and WMH progression, especially in underserved areas where pMRI has huge potential. Here we present a method that segments white matter hyperintensity and 36 brain regions from scans of any resolution and contrast (including pMRI) without retraining. We show results on eight public datasets and on a private dataset with paired high- and low-field scans (3T and 64mT), where we attain strong correlation between the WMH ($ρ$=.85) and hippocampal volumes (r=.89) estimated at both fields. Our method is publicly available as part of FreeSurfer, at: http://surfer.nmr.mgh.harvard.edu/fswiki/WMH-SynthSeg.

eess.IV↗

arXiv2024

CAST: Clustering Self-Attention using Surrogate Tokens for Efficient Transformers

The Transformer architecture has shown to be a powerful tool for a wide range of tasks. It is based on the self-attention mechanism, which is an inherently computationally expensive operation with quadratic computational complexity: memory usage and compute time increase quadratically with the length of the input sequences, thus limiting the application of Transformers. In this work, we propose a novel Clustering self-Attention mechanism using Surrogate Tokens (CAST), to optimize the attention computation and achieve efficient transformers. CAST utilizes learnable surrogate tokens to construct a cluster affinity matrix, used to cluster the input sequence and generate novel cluster summaries. The self-attention from within each cluster is then combined with the cluster summaries of other clusters, enabling information flow across the entire input sequence. CAST improves efficiency by reducing the complexity from $O(N^2)$ to $O(αN)$ where N is the sequence length, and α is constant according to the number of clusters and samples per cluster. We show that CAST performs better than or comparable to the baseline Transformers on long-range sequence modeling tasks, while also achieving higher results on time and memory efficiency than other efficient transformers.

cs.LG↗

arXiv2024

Regressing Transformers for Data-efficient Visual Place Recognition

Visual place recognition is a critical task in computer vision, especially for localization and navigation systems. Existing methods often rely on contrastive learning: image descriptors are trained to have small distance for similar images and larger distance for dissimilar ones in a latent space. However, this approach struggles to ensure accurate distance-based image similarity representation, particularly when training with binary pairwise labels, and complex re-ranking strategies are required. This work introduces a fresh perspective by framing place recognition as a regression problem, using camera field-of-view overlap as similarity ground truth for learning. By optimizing image descriptors to align directly with graded similarity labels, this approach enhances ranking capabilities without expensive re-ranking, offering data-efficient training and strong generalization across several benchmark datasets.

cs.CV↗

arXiv2023

What do neural networks learn in image classification? A frequency shortcut perspective

Frequency analysis is useful for understanding the mechanisms of representation learning in neural networks (NNs). Most research in this area focuses on the learning dynamics of NNs for regression tasks, while little for classification. This study empirically investigates the latter and expands the understanding of frequency shortcuts. First, we perform experiments on synthetic datasets, designed to have a bias in different frequency bands. Our results demonstrate that NNs tend to find simple solutions for classification, and what they learn first during training depends on the most distinctive frequency characteristics, which can be either low- or high-frequencies. Second, we confirm this phenomenon on natural images. We propose a metric to measure class-wise frequency characteristics and a method to identify frequency shortcuts. The results show that frequency shortcuts can be texture-based or shape-based, depending on what best simplifies the objective. Third, we validate the transferability of frequency shortcuts on out-of-distribution (OOD) test sets. Our results suggest that frequency shortcuts can be transferred across datasets and cannot be fully avoided by larger model capacity and data augmentation. We recommend that future research should focus on effective training schemes mitigating frequency shortcut learning.

cs.LG↗

arXiv2023

DFM-X: Augmentation by Leveraging Prior Knowledge of Shortcut Learning

Neural networks are prone to learn easy solutions from superficial statistics in the data, namely shortcut learning, which impairs generalization and robustness of models. We propose a data augmentation strategy, named DFM-X, that leverages knowledge about frequency shortcuts, encoded in Dominant Frequencies Maps computed for image classification models. We randomly select X% training images of certain classes for augmentation, and process them by retaining the frequencies included in the DFMs of other classes. This strategy compels the models to leverage a broader range of frequencies for classification, rather than relying on specific frequency sets. Thus, the models learn more deep and task-related semantics compared to their counterpart trained with standard setups. Unlike other commonly used augmentation techniques which focus on increasing the visual variations of training data, our method targets exploiting the original data efficiently, by distilling prior knowledge about destructive learning behavior of models from data. Our experimental results demonstrate that DFM-X improves robustness against common corruptions and adversarial attacks. It can be seamlessly integrated with other augmentation techniques to further enhance the robustness of models.

cs.CV↗

arXiv2023

Defocus Blur Synthesis and Deblurring via Interpolation and Extrapolation in Latent Space

Though modern microscopes have an autofocusing system to ensure optimal focus, out-of-focus images can still occur when cells within the medium are not all in the same focal plane, affecting the image quality for medical diagnosis and analysis of diseases. We propose a method that can deblur images as well as synthesize defocus blur. We train autoencoders with implicit and explicit regularization techniques to enforce linearity relations among the representations of different blur levels in the latent space. This allows for the exploration of different blur levels of an object by linearly interpolating/extrapolating the latent representations of images taken at different focal planes. Compared to existing works, we use a simple architecture to synthesize images with flexible blur levels, leveraging the linear latent space. Our regularized autoencoders can effectively mimic blur and deblur, increasing data variety as a data augmentation technique and improving the quality of microscopic images, which would be beneficial for further processing and analysis.

eess.IV↗

arXiv2023

Generalized Contrastive Optimization of Siamese Networks for Place Recognition

Visual place recognition is a challenging task in computer vision and a key component of camera-based localization and navigation systems. Recently, Convolutional Neural Networks (CNNs) achieved high results and good generalization capabilities. They are usually trained using pairs or triplets of images labeled as either similar or dissimilar, in a binary fashion. In practice, the similarity between two images is not binary, but continuous. Furthermore, training these CNNs is computationally complex and involves costly pair and triplet mining strategies. We propose a Generalized Contrastive loss (GCL) function that relies on image similarity as a continuous measure, and use it to train a siamese CNN. Furthermore, we present three techniques for automatic annotation of image pairs with labels indicating their degree of similarity, and deploy them to re-annotate the MSLS, TB-Places, and 7Scenes datasets. We demonstrate that siamese CNNs trained using the GCL function and the improved annotations consistently outperform their binary counterparts. Our models trained on MSLS outperform the state-of-the-art methods, including NetVLAD, NetVLAD-SARE, AP-GeM and Patch-NetVLAD, and generalize well on the Pittsburgh30k, Tokyo 24/7, RobotCar Seasons v2 and Extended CMU Seasons datasets. Furthermore, training a siamese network using the GCL function does not require complex pair mining. We release the source code at https://github.com/marialeyvallina/generalized_contrastive_loss.

cs.CV↗

arXiv2023

Data-efficient Large Scale Place Recognition with Graded Similarity Supervision

Visual place recognition (VPR) is a fundamental task of computer vision for visual localization. Existing methods are trained using image pairs that either depict the same place or not. Such a binary indication does not consider continuous relations of similarity between images of the same place taken from different positions, determined by the continuous nature of camera pose. The binary similarity induces a noisy supervision signal into the training of VPR methods, which stall in local minima and require expensive hard mining algorithms to guarantee convergence. Motivated by the fact that two images of the same place only partially share visual cues due to camera pose differences, we deploy an automatic re-annotation strategy to re-label VPR datasets. We compute graded similarity labels for image pairs based on available localization metadata. Furthermore, we propose a new Generalized Contrastive Loss (GCL) that uses graded similarity labels for training contrastive networks. We demonstrate that the use of the new labels and GCL allow to dispense from hard-pair mining, and to train image descriptors that perform better in VPR by nearest neighbor search, obtaining superior or comparable results than methods that require expensive hard-pair mining and re-ranking techniques. Code and models available at: https://github.com/marialeyvallina/generalized_contrastive_loss

cs.CV↗