Deep Learning approaches in dermatological image classification have shown promising results, yet the field faces significant methodological challenges that impede proper evaluation. This paper presents a dual contribution: first, a systematic analysis of current methodological practices in skin disease classification research, revealing substantial inconsistencies in data preparation, augmentation strategies, and performance reporting; second, a comprehensive training and evaluation framework demonstrated through experiments with the DINOv2-Large vision transformer across three benchmark datasets (HAM10000, DermNet, ISIC Atlas). The analysis identifies concerning patterns, including pre-split data augmentation and validation-based reporting, potentially leading to overestimated metrics, while highlighting the lack of unified methodology standards. The experimental results demonstrate DINOv2's performance in skin disease classification, achieving macro-averaged F1-scores of 0.85 (HAM10000), 0.71 (DermNet), and 0.84 (ISIC Atlas). Attention map analysis reveals critical patterns in the model's decision-making, showing sophisticated feature recognition in typical presentations but significant vulnerabilities with atypical cases and composite images. Our findings highlight the need for standardized evaluation protocols and careful implementation strategies in clinical settings. We propose comprehensive methodological recommendations for model development, evaluation, and clinical deployment, emphasizing rigorous data preparation, systematic error analysis, and specialized protocols for different image types. To promote reproducibility, we provide our implementation code through GitHub. This work establishes a foundation for rigorous evaluation standards in dermatological image classification and provides insights for responsible AI implementation in clinical dermatology.
Hyperspectral image classification presents challenges due to spectral redundancy and complex spatial-spectral dependencies. This paper proposes a novel framework, DCT-Mamba3D, for hyperspectral image classification. DCT-Mamba3D incorporates: (1) a 3D spectral-spatial decorrelation module that applies 3D discrete cosine transform basis functions to reduce both spectral and spatial redundancy, enhancing feature clarity across dimensions; (2) a 3D-Mamba module that leverages a bidirectional state-space model to capture intricate spatial-spectral dependencies; and (3) a global residual enhancement module that stabilizes feature representation, improving robustness and convergence. Extensive experiments on benchmark datasets show that our DCT-Mamba3D outperforms the state-of-the-art methods in challenging scenarios such as the same object in different spectra and different objects in the same spectra.
Which functions can be used as activations in deep neural networks? This article explores families of functions based on orthonormal bases, including the Hermite polynomial basis and the Fourier trigonometric basis, as well as a basis resulting from the tropicalization of a polynomial basis. Our study shows that, through simple variance-preserving initialization and without additional clamping mechanisms, these activations can successfully be used to train deep models, such as GPT-2 for next-token prediction on OpenWebText and ConvNeXt for image classification on ImageNet. Our work addresses the issue of exploding and vanishing activations and gradients, particularly prevalent with polynomial activations, and opens the door for improving the efficiency of large-scale learning tasks. Furthermore, our approach provides insight into the structure of neural networks, revealing that networks with polynomial activations can be interpreted as multivariate polynomial mappings. Finally, using Hermite interpolation, we show that our activations can closely approximate classical ones in pre-trained models by matching both the function and its derivative, making them especially useful for fine-tuning tasks. These activations are available in the torchortho library, which can be accessed via: https://github.com/K-H-Ismail/torchortho.
Measuring the connectivity of water in rivers and streams is essential for effective water resource management. Increased extreme weather events associated with climate change can result in alterations to river and stream connectivity. While traditional stream flow gauges are costly to deploy and limited to large river bodies, trail camera methods are a low-cost and easily deployed alternative to collect hourly data. Image capturing, however requires stream ecologists to manually curate (select and label) tens of thousands of images per year. To improve this workflow, we developed an automated instream trail camera image classification system consisting of three parts: (1) image processing, (2) image augmentation and (3) machine learning. The image preprocessing consists of seven image quality filters, foliage-based luma variance reduction, resizing and bottom-center cropping. Images are balanced using variable amount of generative augmentation using diffusion models and then passed to a machine learning classification model in labeled form. By using the vision transformer architecture and temporal image enhancement in our framework, we are able to increase the 75% base accuracy to 90% for a new unseen site image. We make use of a dataset captured and labeled by staff from the Connecticut Department of Energy and Environmental Protection between 2018-2020. Our results indicate that a combination of temporal image processing and attention-based models are effective at classifying unseen river connectivity images.
With the rapid development of modern transportation systems and the exponential growth of logistics volumes, intelligent X-ray-based security inspection systems play a crucial role in public safety. Although single-view X-ray baggage scanner is widely deployed, they struggles to accurately identify contraband in complex stacking scenarios due to strong viewpoint dependency and inadequate feature representation. To address this, we propose a Dual-View Attention-Guided Network for Efficient X-ray Security Inspection (DAGNet). This study builds on a shared-weight backbone network as the foundation and constructs three key modules that work together: (1) Frequency Domain Interaction Module (FDIM) dynamically enhances features by adjusting frequency components based on inter-view relationships; (2) Dual-View Hierarchical Enhancement Module (DVHEM) employs cross-attention to align features between views and capture hierarchical associations; (3) Convolutional Guided Fusion Module (CGFM) fuses features to suppress redundancy while retaining critical discriminative information. Collectively, these modules substantially improve the performance of dual-view X-ray security inspection. Experimental results demonstrate that DAGNet outperforms existing state-of-the-art approaches across multiple backbone architectures. The code is available at:https://github.com/ShilongHong/DAGNet.
The availability of foundational models (FMs) pre-trained on large-scale data has advanced the state-of-the-art in many computer vision tasks. While FMs have demonstrated good zero-shot performance on many image classification tasks, there is often scope for performance improvement by adapting the FM to the downstream task. However, the data that is required for this adaptation typically exists in silos across multiple entities (data owners) and cannot be collated at a central location due to regulations and privacy concerns. At the same time, a learning service provider (LSP) who owns the FM cannot share the model with the data owners due to proprietary reasons. In some cases, the data owners may not even have the resources to store such large FMs. Hence, there is a need for algorithms to adapt the FM in a double-blind federated manner, i.e., the data owners do not know the FM or each other's data, and the LSP does not see the data for the downstream tasks. In this work, we propose a framework for double-blind federated adaptation of FMs using fully homomorphic encryption (FHE). The proposed framework first decomposes the FM into a sequence of FHE-friendly blocks through knowledge distillation. The resulting FHE-friendly model is adapted for the downstream task via low-rank parallel adapters that can be learned without backpropagation through the FM. Since the proposed framework requires the LSP to share intermediate representations with the data owners, we design a privacy-preserving permutation scheme to prevent the data owners from learning the FM through model extraction attacks. Finally, a secure aggregation protocol is employed for federated learning of the low-rank parallel adapters. Empirical results on four datasets demonstrate the practical feasibility of the proposed framework.
Although backpropagation is widely accepted as a training algorithm for artificial neural networks, researchers are always looking for inspiration from the brain to find ways with potentially better performance. Forward-Forward is a new training algorithm that is more similar to what occurs in the brain, although there is a significant performance gap compared to backpropagation. In the Forward-Forward algorithm, the loss functions are placed after each layer, and the updating of a layer is done using two local forward passes and one local backward pass. Forward-Forward is in its early stages and has been designed and evaluated on simple multi-layer perceptron networks to solve image classification tasks. In this work, we have extended the use of this algorithm to a more complex and modern network, namely the Vision Transformer. Inspired by insights from contrastive learning, we have attempted to revise this algorithm, leading to the introduction of Contrastive Forward-Forward. Experimental results show that our proposed algorithm performs significantly better than the baseline Forward-Forward leading to an increase of up to 10% in accuracy and boosting the convergence speed by 5 to 20 times on Vision Transformer. Furthermore, if we take Cross Entropy as the baseline loss function in backpropagation, it will be demonstrated that the proposed modifications to the baseline Forward-Forward reduce its performance gap compared to backpropagation on Vision Transformer, and even outperforms it in certain conditions, such as inaccurate supervision.
We present an all-digital programmable machine learning accelerator chip for image classification, underpinning on the Tsetlin machine (TM) principles. The TM is an emerging machine learning algorithm founded on propositional logic, utilizing sub-pattern recognition expressions called clauses. The accelerator implements the coalesced TM version with convolution, and classifies booleanized images of 28$\times$28 pixels with 10 categories. A configuration with 128 clauses is used in a highly parallel architecture. Fast clause evaluation is achieved by keeping all clause weights and Tsetlin automata (TA) action signals in registers. The chip is implemented in a 65 nm low-leakage CMOS technology, and occupies an active area of 2.7 mm$^2$. At a clock frequency of 27.8 MHz, the accelerator achieves 60.3k classifications per second, and consumes 8.6 nJ per classification. This demonstrates the energy-efficiency of the TM, which was the main motivation for developing this chip. The latency for classifying a single image is 25.4 $\mu$s which includes system timing overhead. The accelerator achieves 97.42%, 84.54% and 82.55% test accuracies for the datasets MNIST, Fashion-MNIST and Kuzushiji-MNIST, respectively, matching the TM software models.
X-ray imaging is pivotal in medical diagnostics, offering non-invasive insights into a range of health conditions. Recently, vision-language models, such as the Contrastive Language-Image Pretraining (CLIP) model, have demonstrated potential in improving diagnostic accuracy by leveraging large-scale image-text datasets. However, since CLIP was not initially designed for medical images, several CLIP-like models trained specifically on medical images have been developed. Despite their enhanced performance, issues of fairness - particularly regarding demographic attributes - remain largely unaddressed. In this study, we perform a comprehensive fairness analysis of CLIP-like models applied to X-ray image classification. We assess their performance and fairness across diverse patient demographics and disease categories using zero-shot inference and various fine-tuning techniques, including Linear Probing, Multilayer Perceptron (MLP), Low-Rank Adaptation (LoRA), and full fine-tuning. Our results indicate that while fine-tuning improves model accuracy, fairness concerns persist, highlighting the need for further fairness interventions in these foundational models.
Machine unlearning seeks to remove the influence of specified data from a trained model. While metrics such as unlearning accuracy (UA) and membership inference attack (MIA) provide baselines for assessing unlearning performance, they fall short of evaluating the forgetting reliability. In this paper, we find that the data misclassified across UA and MIA still have their ground truth labels included in the prediction set from the uncertainty quantification perspective, which raises a fake unlearning issue. To address this issue, we propose two novel metrics inspired by conformal prediction that more reliably evaluate forgetting quality. Building on these insights, we further propose a conformal prediction-based unlearning framework that integrates conformal prediction into Carlini & Wagner adversarial attack loss, which can significantly push the ground truth label out of the conformal prediction set. Through extensive experiments on image classification task, we demonstrate both the effectiveness of our proposed metrics and the superiority of our unlearning framework, which improves the UA of existing unlearning methods by an average of 6.6% through the incorporation of a tailored loss term alone.
Transformers have achieved state-of-the-art performance in numerous tasks. In this paper, we propose a continuous-time formulation of transformers. Specifically, we consider a dynamical system whose governing equation is parametrized by transformer blocks. We leverage optimal transport theory to regularize the training problem, which enhances stability in training and improves generalization of the resulting model. Moreover, we demonstrate in theory that this regularization is necessary as it promotes uniqueness and regularity of solutions. Our model is flexible in that almost any existing transformer architectures can be adopted to construct the dynamical system with only slight modifications to the existing code. We perform extensive numerical experiments on tasks motivated by natural language processing, image classification, and point cloud classification. Our experimental results show that the proposed method improves the performance of its discrete counterpart and outperforms relevant comparing models.
Spiking neural networks (SNNs) are bio-inspired networks that mimic how neurons in the brain communicate through discrete spikes, which have great potential in various tasks due to their energy efficiency and temporal processing capabilities. SNNs with self-attention mechanisms (spiking Transformers) have recently shown great advancements in various tasks, and inspired by traditional Transformers, several studies have demonstrated that spiking absolute positional encoding can help capture sequential relationships for input data, enhancing the capabilities of spiking Transformers for tasks such as sequential modeling and image classification. However, how to incorporate relative positional information into SNNs remains a challenge. In this paper, we introduce several strategies to approximate relative positional encoding (RPE) in spiking Transformers while preserving the binary nature of spikes. Firstly, we formally prove that encoding relative distances with Gray Code ensures that the binary representations of positional indices maintain a constant Hamming distance whenever their decimal values differ by a power of two, and we propose Gray-PE based on this property. In addition, we propose another RPE method called Log-PE, which combines the logarithmic form of the relative distance matrix directly into the spiking attention map. Furthermore, we extend our RPE methods to a two-dimensional form, making them suitable for processing image patches. We evaluate our RPE methods on various tasks, including time series forecasting, text classification, and patch-based image classification, and the experimental results demonstrate a satisfying performance gain by incorporating our RPE methods across many architectures.
Rare events, due to their infrequent occurrences, do not have much data, and hence deep learning techniques fail in estimating the distribution for such data. Open-vocabulary models represent an innovative approach to image classification. Unlike traditional models, these models classify images into any set of categories specified with natural language prompts during inference. These prompts usually comprise manually crafted templates (e.g., 'a photo of a {}') that are filled in with the names of each category. This paper introduces a simple yet effective method for generating highly accurate and contextually descriptive prompts containing discriminative characteristics. Rare event detection, especially in medicine, is more challenging due to low inter-class and high intra-class variability. To address these, we propose a novel approach that uses domain-specific expert knowledge on rare events to generate customized and contextually relevant prompts, which are then used by large language models for image classification. Our zero-shot, privacy-preserving method enhances rare event classification without additional training, outperforming state-of-the-art techniques.
Despite the significant success of deep learning models in computer vision, they often exhibit systematic failures on specific data subsets, known as error slices. Identifying and mitigating these error slices is crucial to enhancing model robustness and reliability in real-world scenarios. In this paper, we introduce DebugAgent, an automated framework for error slice discovery and model repair. DebugAgent first generates task-specific visual attributes to highlight instances prone to errors through an interpretable and structured process. It then employs an efficient slice enumeration algorithm to systematically identify error slices, overcoming the combinatorial challenges that arise during slice exploration. Additionally, DebugAgent extends its capabilities by predicting error slices beyond the validation set, addressing a key limitation of prior approaches. Extensive experiments across multiple domains, including image classification, pose estimation, and object detection - show that DebugAgent not only improves the coherence and precision of identified error slices but also significantly enhances the model repair capabilities.
There has been significant focus on creating neuro-symbolic models for interpretable image classification using Convolutional Neural Networks (CNNs). These methods aim to replace the CNN with a neuro-symbolic model consisting of the CNN, which is used as a feature extractor, and an interpretable rule-set extracted from the CNN itself. While these approaches provide interpretability through the extracted rule-set, they often compromise accuracy compared to the original CNN model. In this paper, we identify the root cause of this accuracy loss as the post-training binarization of filter activations to extract the rule-set. To address this, we propose a novel sparsity loss function that enables class-specific filter binarization during CNN training, thus minimizing information loss when extracting the rule-set. We evaluate several training strategies with our novel sparsity loss, analyzing their effectiveness and providing guidance on their appropriate use. Notably, we set a new benchmark, achieving a 9% improvement in accuracy and a 53% reduction in rule-set size on average, compared to the previous SOTA, while coming within 3% of the original CNN's accuracy. This highlights the significant potential of interpretable neuro-symbolic models as viable alternatives to black-box CNNs.
The development of deep learning techniques is a leading field applied to cases in which medical data is used, particularly in cases of image diagnosis. This type of data has privacy and legal restrictions that in many cases prevent it from being processed from central servers. However, in this area collaboration between different research centers, in order to create models as robust as possible, trained with the largest quantity and diversity of data available, is a critical point to be taken into account. In this sense, the application of privacy aware distributed architectures, such as federated learning arises. When applying this type of architecture, the server aggregates the different local models trained with the data of each data owner to build a global model. This point is critical and therefore it is fundamental to analyze different ways of aggregation according to the use case, taking into account the distribution of the clients, the characteristics of the model, etc. In this paper we propose a novel aggregation strategy and we apply it to a use case of cerebral magnetic resonance image classification. In this use case the aggregation function proposed manages to improve the convergence obtained over the rounds of the federated learning process in relation to different aggregation strategies classically implemented and applied.
Hyperspectral image (HSI) classification aims at categorizing each pixel in an HSI into a specific land cover class, which is crucial for applications like remote sensing, environmental monitoring, and agriculture. Although deep learning-based HSI classification methods have achieved significant advancements, existing methods still rely on manually labeled data for training, which is both time-consuming and labor-intensive. To address this limitation, we introduce a novel zero-shot hyperspectral image classification framework based on CLIP (SPECIAL), aiming to eliminate the need for manual annotations. The SPECIAL framework consists of two main stages: (1) CLIP-based pseudo-label generation, and (2) noisy label learning. In the first stage, HSI is spectrally interpolated to produce RGB bands. These bands are subsequently classified using CLIP, resulting in noisy pseudo-labels that are accompanied by confidence scores. To improve the quality of these labels, we propose a scaling strategy that fuses predictions from multiple spatial scales. In the second stage, spectral information and a label refinement technique are incorporated to mitigate label noise and further enhance classification accuracy. Experimental results on three benchmark datasets demonstrate that our SPECIAL outperforms existing methods in zero-shot HSI classification, showing its potential for more practical applications. The code is available at https://github.com/LiPang/SPECIAL.
Source-free domain adaptation in visual emotion recognition (SFDA-VER) is a highly challenging task that requires adapting VER models to the target domain without relying on source data, which is of great significance for data privacy protection. However, due to the unignorable disparities between visual emotion data and traditional image classification data, existing SFDA methods perform poorly on this task. In this paper, we investigate the SFDA-VER task from a fuzzy perspective and identify two key issues: fuzzy emotion labels and fuzzy pseudo-labels. These issues arise from the inherent uncertainty of emotion annotations and the potential mispredictions in pseudo-labels. To address these issues, we propose a novel fuzzy-aware loss (FAL) to enable the VER model to better learn and adapt to new domains under fuzzy labels. Specifically, FAL modifies the standard cross entropy loss and focuses on adjusting the losses of non-predicted categories, which prevents a large number of uncertain or incorrect predictions from overwhelming the VER model during adaptation. In addition, we provide a theoretical analysis of FAL and prove its robustness in handling the noise in generated pseudo-labels. Extensive experiments on 26 domain adaptation sub-tasks across three benchmark datasets demonstrate the effectiveness of our method.
This paper aims to improve the robustness of a small global model while maintaining clean accuracy under adversarial attacks and non-IID challenges in federated learning. By leveraging the concise knowledge embedded in the class probabilities from a pre-trained model for both clean and adversarial image classification, we propose a Pre-trained Model-guided Adversarial Federated Learning (PM-AFL) training paradigm. This paradigm integrates vanilla mixture and adversarial mixture knowledge distillation to effectively balance accuracy and robustness while promoting local models to learn from diverse data. Specifically, for clean accuracy, we adopt a dual distillation strategy where the class probabilities of randomly paired images and their blended versions are aligned between the teacher model and the local models. For adversarial robustness, we use a similar distillation approach but replace clean samples on the local side with adversarial examples. Moreover, considering the bias between local and global models, we also incorporate a consistency regularization term to ensure that local adversarial predictions stay aligned with their corresponding global clean ones. These strategies collectively enable local models to absorb diverse knowledge from the teacher model while maintaining close alignment with the global model, thereby mitigating overfitting to local optima and enhancing the generalization of the global model. Experiments demonstrate that the PM-AFL-based paradigm outperforms other methods that integrate defense strategies by a notable margin.
We introduce Feasible Learning (FL), a sample-centric learning paradigm where models are trained by solving a feasibility problem that bounds the loss for each training sample. In contrast to the ubiquitous Empirical Risk Minimization (ERM) framework, which optimizes for average performance, FL demands satisfactory performance on every individual data point. Since any model that meets the prescribed performance threshold is a valid FL solution, the choice of optimization algorithm and its dynamics play a crucial role in shaping the properties of the resulting solutions. In particular, we study a primal-dual approach which dynamically re-weights the importance of each sample during training. To address the challenge of setting a meaningful threshold in practice, we introduce a relaxation of FL that incorporates slack variables of minimal norm. Our empirical analysis, spanning image classification, age regression, and preference optimization in large language models, demonstrates that models trained via FL can learn from data while displaying improved tail behavior compared to ERM, with only a marginal impact on average performance.