CLIP showcases exceptional cross-modal matching capabilities due to its training on image-text contrastive learning tasks. However, without specific optimization for unimodal scenarios, its performance in single-modality feature extraction might be suboptimal. Despite this, some studies have directly used CLIP's image encoder for tasks like few-shot classification, introducing a misalignment between its pre-training objectives and feature extraction methods. This inconsistency can diminish the quality of the image's feature representation, adversely affecting CLIP's effectiveness in target tasks. In this paper, we view text features as precise neighbors of image features in CLIP's space and present a novel CrOss-moDal nEighbor Representation(CODER) based on the distance structure between images and their neighbor texts. This feature extraction method aligns better with CLIP's pre-training objectives, thereby fully leveraging CLIP's robust cross-modal capabilities. The key to construct a high-quality CODER lies in how to create a vast amount of high-quality and diverse texts to match with images. We introduce the Auto Text Generator(ATG) to automatically generate the required texts in a data-free and training-free manner. We apply CODER to CLIP's zero-shot and few-shot image classification tasks. Experiment results across various datasets and models confirm CODER's effectiveness. Code is available at:https://github.com/YCaigogogo/CVPR24-CODER.
Land cover analysis using hyperspectral images (HSI) remains an open problem due to their low spatial resolution and complex spectral information. Recent studies are primarily dedicated to designing Transformer-based architectures for spatial-spectral long-range dependencies modeling, which is computationally expensive with quadratic complexity. Selective structured state space model (Mamba), which is efficient for modeling long-range dependencies with linear complexity, has recently shown promising progress. However, its potential in hyperspectral image processing that requires handling numerous spectral bands has not yet been explored. In this paper, we innovatively propose S$^2$Mamba, a spatial-spectral state space model for hyperspectral image classification, to excavate spatial-spectral contextual features, resulting in more efficient and accurate land cover analysis. In S$^2$Mamba, two selective structured state space models through different dimensions are designed for feature extraction, one for spatial, and the other for spectral, along with a spatial-spectral mixture gate for optimal fusion. More specifically, S$^2$Mamba first captures spatial contextual relations by interacting each pixel with its adjacent through a Patch Cross Scanning module and then explores semantic information from continuous spectral bands through a Bi-directional Spectral Scanning module. Considering the distinct expertise of the two attributes in homogenous and complicated texture scenes, we realize the Spatial-spectral Mixture Gate by a group of learnable matrices, allowing for the adaptive incorporation of representations learned across different dimensions. Extensive experiments conducted on HSI classification benchmarks demonstrate the superiority and prospect of S$^2$Mamba. The code will be made available at: https://github.com/PURE-melo/S2Mamba.
Pooling layers (e.g., max and average) may overlook important information encoded in the spatial arrangement of pixel intensity and/or feature values. We propose a novel lacunarity pooling layer that aims to capture the spatial heterogeneity of the feature maps by evaluating the variability within local windows. The layer operates at multiple scales, allowing the network to adaptively learn hierarchical features. The lacunarity pooling layer can be seamlessly integrated into any artificial neural network architecture. Experimental results demonstrate the layer's effectiveness in capturing intricate spatial patterns, leading to improved feature extraction capabilities. The proposed approach holds promise in various domains, especially in agricultural image analysis tasks. This work contributes to the evolving landscape of artificial neural network architectures by introducing a novel pooling layer that enriches the representation of spatial features. Our code is publicly available.
The recent prevalence of publicly accessible, large medical imaging datasets has led to a proliferation of artificial intelligence (AI) models for cardiovascular image classification and analysis. At the same time, the potentially significant impacts of these models have motivated the development of a range of explainable AI (XAI) methods that aim to explain model predictions given certain image inputs. However, many of these methods are not developed or evaluated with domain experts, and explanations are not contextualized in terms of medical expertise or domain knowledge. In this paper, we propose a novel framework and python library, MiMICRI, that provides domain-centered counterfactual explanations of cardiovascular image classification models. MiMICRI helps users interactively select and replace segments of medical images that correspond to morphological structures. From the counterfactuals generated, users can then assess the influence of each segment on model predictions, and validate the model against known medical facts. We evaluate this library with two medical experts. Our evaluation demonstrates that a domain-centered XAI approach can enhance the interpretability of model explanations, and help experts reason about models in terms of relevant domain knowledge. However, concerns were also surfaced about the clinical plausibility of the counterfactuals generated. We conclude with a discussion on the generalizability and trustworthiness of the MiMICRI framework, as well as the implications of our findings on the development of domain-centered XAI methods for model interpretability in healthcare contexts.
Unsupervised domain adaptation (UDA) aims to transfer knowledge from a labeled source domain to an unlabeled target domain. The most recent UDA methods always resort to adversarial training to yield state-of-the-art results and a dominant number of existing UDA methods employ convolutional neural networks (CNNs) as feature extractors to learn domain invariant features. Vision transformer (ViT) has attracted tremendous attention since its emergence and has been widely used in various computer vision tasks, such as image classification, object detection, and semantic segmentation, yet its potential in adversarial domain adaptation has never been investigated. In this paper, we fill this gap by employing the ViT as the feature extractor in adversarial domain adaptation. Moreover, we empirically demonstrate that ViT can be a plug-and-play component in adversarial domain adaptation, which means directly replacing the CNN-based feature extractor in existing UDA methods with the ViT-based feature extractor can easily obtain performance improvement. The code is available at https://github.com/LluckyYH/VT-ADA.
The integration of deep learning based systems in clinical practice is often impeded by challenges rooted in limited and heterogeneous medical datasets. In addition, the field has increasingly prioritized marginal performance gains on a few, narrowly scoped benchmarks over clinical applicability, slowing down meaningful algorithmic progress. This trend often results in excessive fine-tuning of existing methods on selected datasets rather than fostering clinically relevant innovations. In response, this work introduces a comprehensive benchmark for the MedMNIST+ dataset collection, designed to diversify the evaluation landscape across several imaging modalities, anatomical regions, classification tasks and sample sizes. We systematically reassess commonly used Convolutional Neural Networks (CNNs) and Vision Transformer (ViT) architectures across distinct medical datasets, training methodologies, and input resolutions to validate and refine existing assumptions about model effectiveness and development. Our findings suggest that computationally efficient training schemes and modern foundation models offer viable alternatives to costly end-to-end training. Additionally, we observe that higher image resolutions do not consistently improve performance beyond a certain threshold. This highlights the potential benefits of using lower resolutions, particularly in prototyping stages, to reduce computational demands without sacrificing accuracy. Notably, our analysis reaffirms the competitiveness of CNNs compared to ViTs, emphasizing the importance of comprehending the intrinsic capabilities of different architectures. Finally, by establishing a standardized evaluation framework, we aim to enhance transparency, reproducibility, and comparability within the MedMNIST+ dataset collection. Code is available at https://github.com/sdoerrich97/rethinking-model-prototyping-MedMNISTPlus .
The success of contrastive language-image pretraining (CLIP) relies on the supervision from the pairing between images and captions, which tends to be noisy in web-crawled data. We present Mixture of Data Experts (MoDE) and learn a system of CLIP data experts via clustering. Each data expert is trained on one data cluster, being less sensitive to false negative noises in other clusters. At inference time, we ensemble their outputs by applying weights determined through the correlation between task metadata and cluster conditions. To estimate the correlation precisely, the samples in one cluster should be semantically similar, but the number of data experts should still be reasonable for training and inference. As such, we consider the ontology in human language and propose to use fine-grained cluster centers to represent each data expert at a coarse-grained level. Experimental studies show that four CLIP data experts on ViT-B/16 outperform the ViT-L/14 by OpenAI CLIP and OpenCLIP on zero-shot image classification but with less ($<$35\%) training cost. Meanwhile, MoDE can train all data expert asynchronously and can flexibly include new data experts. The code is available at https://github.com/facebookresearch/MetaCLIP/tree/main/mode.
Hyperspectral Image Classification (HSC) presents significant challenges owing to the high dimensionality and intricate nature of Hyperspectral (HS) data. While traditional Machine Learning (TML) approaches have demonstrated effectiveness, they often encounter substantial obstacles in real-world applications, including the variability of optimal feature sets, subjectivity in human-driven design, inherent biases, and methodological limitations. Specifically, TML suffers from the curse of dimensionality, difficulties in feature selection and extraction, insufficient consideration of spatial information, limited robustness against noise, scalability issues, and inadequate adaptability to complex data distributions. In recent years, Deep Learning (DL) techniques have emerged as robust solutions to address these challenges. This survey offers a comprehensive overview of current trends and future prospects in HSC, emphasizing advancements from DL models to the increasing adoption of Transformer and Mamba Model architectures. We systematically review key concepts, methodologies, and state-of-the-art approaches in DL for HSC. Furthermore, we investigate the potential of Transformer-based models and the Mamba Model in HSC, detailing their advantages and challenges. Emerging trends in HSC are explored, including in-depth discussions on Explainable AI and Interoperability concepts, alongside Diffusion Models for image denoising, feature extraction, and image fusion. Comprehensive experimental results were conducted on three HS datasets to substantiate the efficacy of various conventional DL models and Transformers. Additionally, we identify several open challenges and pertinent research questions in the field of HSC. Finally, we outline future research directions and potential applications aimed at enhancing the accuracy and efficiency of HSC.
The traditional Transformer model encounters challenges with variable-length input sequences, particularly in Hyperspectral Image Classification (HSIC), leading to efficiency and scalability concerns. To overcome this, we propose a pyramid-based hierarchical transformer (PyFormer). This innovative approach organizes input data hierarchically into segments, each representing distinct abstraction levels, thereby enhancing processing efficiency for lengthy sequences. At each level, a dedicated transformer module is applied, effectively capturing both local and global context. Spatial and spectral information flow within the hierarchy facilitates communication and abstraction propagation. Integration of outputs from different levels culminates in the final input representation. Experimental results underscore the superiority of the proposed method over traditional approaches. Additionally, the incorporation of disjoint samples augments robustness and reliability, thereby highlighting the potential of our approach in advancing HSIC. The source code is available at https://github.com/mahmad00/PyFormer.
Disjoint sampling is critical for rigorous and unbiased evaluation of state-of-the-art (SOTA) models. When training, validation, and test sets overlap or share data, it introduces a bias that inflates performance metrics and prevents accurate assessment of a model's true ability to generalize to new examples. This paper presents an innovative disjoint sampling approach for training SOTA models on Hyperspectral image classification (HSIC) tasks. By separating training, validation, and test data without overlap, the proposed method facilitates a fairer evaluation of how well a model can classify pixels it was not exposed to during training or validation. Experiments demonstrate the approach significantly improves a model's generalization compared to alternatives that include training and validation data in test data. By eliminating data leakage between sets, disjoint sampling provides reliable metrics for benchmarking progress in HSIC. Researchers can have confidence that reported performance truly reflects a model's capabilities for classifying new scenes, not just memorized pixels. This rigorous methodology is critical for advancing SOTA models and their real-world application to large-scale land mapping with Hyperspectral sensors. The source code is available at https://github.com/mahmad00/Disjoint-Sampling-for-Hyperspectral-Image-Classification.
Mounting evidence in explainability for artificial intelligence (XAI) research suggests that good explanations should be tailored to individual tasks and should relate to concepts relevant to the task. However, building task specific explanations is time consuming and requires domain expertise which can be difficult to integrate into generic XAI methods. A promising approach towards designing useful task specific explanations with domain experts is based on compositionality of semantic concepts. Here, we present a novel approach that enables domain experts to quickly create concept-based explanations for computer vision tasks intuitively via natural language. Leveraging recent progress in deep generative methods we propose to generate visual concept-based prototypes via text-to-image methods. These prototypes are then used to explain predictions of computer vision models via a simple k-Nearest-Neighbors routine. The modular design of CoProNN is simple to implement, it is straightforward to adapt to novel tasks and allows for replacing the classification and text-to-image models as more powerful models are released. The approach can be evaluated offline against the ground-truth of predefined prototypes that can be easily communicated also to domain experts as they are based on visual concepts. We show that our strategy competes very well with other concept-based XAI approaches on coarse grained image classification tasks and may even outperform those methods on more demanding fine grained tasks. We demonstrate the effectiveness of our method for human-machine collaboration settings in qualitative and quantitative user studies. All code and experimental data can be found in our GitHub $\href{https://github.com/TeodorChiaburu/beexplainable}{repository}$.
In recent years, Vision Transformers (ViTs) have shown promising classification performance over Convolutional Neural Networks (CNNs) due to their self-attention mechanism. Many researchers have incorporated ViTs for Hyperspectral Image (HSI) classification. HSIs are characterised by narrow contiguous spectral bands, providing rich spectral data. Although ViTs excel with sequential data, they cannot extract spectral-spatial information like CNNs. Furthermore, to have high classification performance, there should be a strong interaction between the HSI token and the class (CLS) token. To solve these issues, we propose a 3D-Convolution guided Spectral-Spatial Transformer (3D-ConvSST) for HSI classification that utilizes a 3D-Convolution Guided Residual Module (CGRM) in-between encoders to "fuse" the local spatial and spectral information and to enhance the feature propagation. Furthermore, we forego the class token and instead apply Global Average Pooling, which effectively encodes more discriminative and pertinent high-level features for classification. Extensive experiments have been conducted on three public HSI datasets to show the superiority of the proposed model over state-of-the-art traditional, convolutional, and Transformer models. The code is available at https://github.com/ShyamVarahagiri/3D-ConvSST.
In this paper, we present a simple yet effective contrastive knowledge distillation approach, which can be formulated as a sample-wise alignment problem with intra- and inter-sample constraints. Unlike traditional knowledge distillation methods that concentrate on maximizing feature similarities or preserving class-wise semantic correlations between teacher and student features, our method attempts to recover the "dark knowledge" by aligning sample-wise teacher and student logits. Specifically, our method first minimizes logit differences within the same sample by considering their numerical values, thus preserving intra-sample similarities. Next, we bridge semantic disparities by leveraging dissimilarities across different samples. Note that constraints on intra-sample similarities and inter-sample dissimilarities can be efficiently and effectively reformulated into a contrastive learning framework with newly designed positive and negative pairs. The positive pair consists of the teacher's and student's logits derived from an identical sample, while the negative pairs are formed by using logits from different samples. With this formulation, our method benefits from the simplicity and efficiency of contrastive learning through the optimization of InfoNCE, yielding a run-time complexity that is far less than $O(n^2)$, where $n$ represents the total number of training samples. Furthermore, our method can eliminate the need for hyperparameter tuning, particularly related to temperature parameters and large batch sizes. We conduct comprehensive experiments on three datasets including CIFAR-100, ImageNet-1K, and MS COCO. Experimental results clearly confirm the effectiveness of the proposed method on both image classification and object detection tasks. Our source codes will be publicly available at https://github.com/wencheng-zhu/CKD.
Neural networks are trained by minimizing a loss function that defines the discrepancy between the predicted model output and the target value. The selection of the loss function is crucial to achieve task-specific behaviour and highly influences the capability of the model. A variety of loss functions have been proposed for a wide range of tasks affecting training and model performance. For classification tasks, the cross entropy is the de-facto standard and usually the first choice. Here, we try to experimentally challenge the well-known loss functions, including cross entropy (CE) loss, by utilizing the genetic programming (GP) approach, a population-based evolutionary algorithm. GP constructs loss functions from a set of operators and leaf nodes and these functions are repeatedly recombined and mutated to find an optimal structure. Experiments were carried out on different small-sized datasets CIFAR-10, CIFAR-100 and Fashion-MNIST using an Inception model. The 5 best functions found were evaluated for different model architectures on a set of standard datasets ranging from 2 to 102 classes and very different sizes. One function, denoted as Next Generation Loss (NGL), clearly stood out showing same or better performance for all tested datasets compared to CE. To evaluate the NGL function on a large-scale dataset, we tested its performance on the Imagenet-1k dataset where it showed improved top-1 accuracy compared to models trained with identical settings and other losses. Finally, the NGL was trained on a segmentation downstream task for Pascal VOC 2012 and COCO-Stuff164k datasets improving the underlying model performance.
Masked image modeling (MIM) pre-training for large-scale vision transformers (ViTs) has enabled promising downstream performance on top of the learned self-supervised ViT features. In this paper, we question if the \textit{extremely simple} lightweight ViTs' fine-tuning performance can also benefit from this pre-training paradigm, which is considerably less studied yet in contrast to the well-established lightweight architecture design methodology. We use an observation-analysis-solution flow for our study. We first systematically observe different behaviors among the evaluated pre-training methods with respect to the downstream fine-tuning data scales. Furthermore, we analyze the layer representation similarities and attention maps across the obtained models, which clearly show the inferior learning of MIM pre-training on higher layers, leading to unsatisfactory transfer performance on data-insufficient downstream tasks. This finding is naturally a guide to designing our distillation strategies during pre-training to solve the above deterioration problem. Extensive experiments have demonstrated the effectiveness of our approach. Our pre-training with distillation on pure lightweight ViTs with vanilla/hierarchical design ($5.7M$/$6.5M$) can achieve $79.4\%$/$78.9\%$ top-1 accuracy on ImageNet-1K. It also enables SOTA performance on the ADE20K segmentation task ($42.8\%$ mIoU) and LaSOT tracking task ($66.1\%$ AUC) in the lightweight regime. The latter even surpasses all the current SOTA lightweight CPU-realtime trackers.
Semi-supervised image classification, leveraging pseudo supervision and consistency regularization, has demonstrated remarkable success. However, the ongoing challenge lies in fully exploiting the potential of unlabeled data. To address this, we employ information entropy neural estimation to utilize the potential of unlabeled samples. Inspired by contrastive learning, the entropy is estimated by maximizing a lower bound on mutual information across different augmented views. Moreover, we theoretically analyze that the information entropy of the posterior of an image classifier is approximated by maximizing the likelihood function of the softmax predictions. Guided by these insights, we optimize our model from both perspectives to ensure that the predicted probability distribution closely aligns with the ground-truth distribution. Given the theoretical connection to information entropy, we name our method InfoMatch. Through extensive experiments, we show its superior performance. The source code is available at https://github.com/kunzhan/InfoMatch.
Large vision-language models revolutionized image classification and semantic segmentation paradigms. However, they typically assume a pre-defined set of categories, or vocabulary, at test time for composing textual prompts. This assumption is impractical in scenarios with unknown or evolving semantic context. Here, we address this issue and introduce the Vocabulary-free Image Classification (VIC) task, which aims to assign a class from an unconstrained language-induced semantic space to an input image without needing a known vocabulary. VIC is challenging due to the vastness of the semantic space, which contains millions of concepts, including fine-grained categories. To address VIC, we propose Category Search from External Databases (CaSED), a training-free method that leverages a pre-trained vision-language model and an external database. CaSED first extracts the set of candidate categories from the most semantically similar captions in the database and then assigns the image to the best-matching candidate category according to the same vision-language model. Furthermore, we demonstrate that CaSED can be applied locally to generate a coarse segmentation mask that classifies image regions, introducing the task of Vocabulary-free Semantic Segmentation. CaSED and its variants outperform other more complex vision-language models, on classification and semantic segmentation benchmarks, while using much fewer parameters.
In this paper, we tackle Multi-Source Domain Adaptation (MSDA), a task in transfer learning where one adapts multiple heterogeneous, labeled source probability measures towards a different, unlabeled target measure. We propose a novel framework for MSDA, based on Optimal Transport (OT) and Gaussian Mixture Models (GMMs). Our framework has two key advantages. First, OT between GMMs can be solved efficiently via linear programming. Second, it provides a convenient model for supervised learning, especially classification, as components in the GMM can be associated with existing classes. Based on the GMM-OT problem, we propose a novel technique for calculating barycenters of GMMs. Based on this novel algorithm, we propose two new strategies for MSDA: GMM-Wasserstein Barycenter Transport (WBT) and GMM-Dataset Dictionary Learning (DaDiL). We empirically evaluate our proposed methods on four benchmarks in image classification and fault diagnosis, showing that we improve over the prior art while being faster and involving fewer parameters. Our code is publicly available at https://github.com/eddardd/gmm_msda
Recent advancements in general-purpose or domain-specific multimodal large language models (LLMs) have witnessed remarkable progress for medical decision-making. However, they are designated for specific classification or generative tasks, and require model training or finetuning on large-scale datasets with sizeable parameters and tremendous computing, hindering their clinical utility across diverse resource-constrained scenarios in practice. In this paper, we propose a novel and lightweight framework Med-MoE (Mixture-of-Experts) that tackles both discriminative and generative multimodal medical tasks. The learning of Med-MoE consists of three steps: multimodal medical alignment, instruction tuning and routing, and domain-specific MoE tuning. After aligning multimodal medical images with LLM tokens, we then enable the model for different multimodal medical tasks with instruction tuning, together with a trainable router tailored for expert selection across input modalities. Finally, the model is tuned by integrating the router with multiple domain-specific experts, which are selectively activated and further empowered by meta expert. Comprehensive experiments on both open- and close-end medical question answering (Med-VQA) and image classification tasks across datasets such as VQA-RAD, SLAKE and Path-VQA demonstrate that our model can achieve performance superior to or on par with state-of-the-art baselines, while only requiring approximately 30\%-50\% of activated model parameters. Extensive analysis and ablations corroborate the effectiveness and practical utility of our method.
Images captured from the real world are often affected by different types of noise, which can significantly impact the performance of Computer Vision systems and the quality of visual data. This study presents a novel approach for defect detection in casting product noisy images, specifically focusing on submersible pump impellers. The methodology involves utilizing deep learning models such as VGG16, InceptionV3, and other models in both the spatial and frequency domains to identify noise types and defect status. The research process begins with preprocessing images, followed by applying denoising techniques tailored to specific noise categories. The goal is to enhance the accuracy and robustness of defect detection by integrating noise detection and denoising into the classification pipeline. The study achieved remarkable results using VGG16 for noise type classification in the frequency domain, achieving an accuracy of over 99%. Removal of salt and pepper noise resulted in an average SSIM of 87.9, while Gaussian noise removal had an average SSIM of 64.0, and periodic noise removal yielded an average SSIM of 81.6. This comprehensive approach showcases the effectiveness of the deep AutoEncoder model and median filter, for denoising strategies in real-world industrial applications. Finally, our study reports significant improvements in binary classification accuracy for defect detection compared to previous methods. For the VGG16 classifier, accuracy increased from 94.6% to 97.0%, demonstrating the effectiveness of the proposed noise detection and denoising approach. Similarly, for the InceptionV3 classifier, accuracy improved from 84.7% to 90.0%, further validating the benefits of integrating noise analysis into the classification pipeline.