PyTorchKR Papers

최신 논문

총 10,165편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2305.02932

Image Captioners Sometimes Tell More Than Images They See

Image captioning, a.k.a. "image-to-text," which generates descriptive text from given images, has been rapidly developing throughout the era of deep learning. To what extent is the information in the original image preserved in the descriptive text generated by an image captioner? To answer that question, we have performed experiments involving the classification of images from descriptive text alone, without referring to the images at all, and compared results with those from standard image-based classifiers. We have evaluate several image captioning models with respect to a disaster image classification task, CrisisNLP, and show that descriptive text classifiers can sometimes achieve higher accuracy than standard image-based classifiers. Further, we show that fusing an image-based classifier with a descriptive text classifier can provide improvement in accuracy.

arXiv:2305.02032

Unsupervised Mutual Transformer Learning for Multi-Gigapixel Whole Slide Image Classification

Classification of gigapixel Whole Slide Images (WSIs) is an important prediction task in the emerging area of computational pathology. There has been a surge of research in deep learning models for WSI classification with clinical applications such as cancer detection or prediction of molecular mutations from WSIs. Most methods require expensive and labor-intensive manual annotations by expert pathologists. Weakly supervised Multiple Instance Learning (MIL) methods have recently demonstrated excellent performance; however, they still require large slide-level labeled training datasets that need a careful inspection of each slide by an expert pathologist. In this work, we propose a fully unsupervised WSI classification algorithm based on mutual transformer learning. Instances from gigapixel WSI (i.e., image patches) are transformed into a latent space and then inverse-transformed to the original space. Using the transformation loss, pseudo-labels are generated and cleaned using a transformer label-cleaner. The proposed transformer-based pseudo-label generation and cleaning modules mutually train each other iteratively in an unsupervised manner. A discriminative learning mechanism is introduced to improve normal versus cancerous instance labeling. In addition to unsupervised classification, we demonstrate the effectiveness of the proposed framework for weak supervision for cancer subtype classification as downstream analysis. Extensive experiments on four publicly available datasets show excellent performance compared to the state-of-the-art methods. We intend to make the source code of our algorithm publicly available soon.

arXiv:2305.00983

Detecting Novelties with Empty Classes

For open world applications, deep neural networks (DNNs) need to be aware of previously unseen data and adaptable to evolving environments. Furthermore, it is desirable to detect and learn novel classes which are not included in the DNNs underlying set of semantic classes in an unsupervised fashion. The method proposed in this article builds upon anomaly detection to retrieve out-of-distribution (OoD) data as candidates for new classes. We thereafter extend the DNN by $k$ empty classes and fine-tune it on the OoD data samples. To this end, we introduce two loss functions, which 1) entice the DNN to assign OoD samples to the empty classes and 2) to minimize the inner-class feature distances between them. Thus, instead of ground truth which contains labels for the different novel classes, the DNN obtains a single OoD label together with a distance matrix, which is computed in advance. We perform several experiments for image classification and semantic segmentation, which demonstrate that a DNN can extend its own semantic space by multiple classes without having access to ground truth.

arXiv:2305.01595

On the Impact of Data Quality on Image Classification Fairness

With the proliferation of algorithmic decision-making, increased scrutiny has been placed on these systems. This paper explores the relationship between the quality of the training data and the overall fairness of the models trained with such data in the context of supervised classification. We measure key fairness metrics across a range of algorithms over multiple image classification datasets that have a varying level of noise in both the labels and the training data itself. We describe noise in the labels as inaccuracies in the labelling of the data in the training set and noise in the data as distortions in the data, also in the training set. By adding noise to the original datasets, we can explore the relationship between the quality of the training data and the fairness of the output of the models trained on that data.

arXiv:2305.01160공식 코드 · Papers with Code

Long-Tailed Recognition by Mutual Information Maximization between Latent Features and Ground-Truth Labels

Although contrastive learning methods have shown prevailing performance on a variety of representation learning tasks, they encounter difficulty when the training dataset is long-tailed. Many researchers have combined contrastive learning and a logit adjustment technique to address this problem, but the combinations are done ad-hoc and a theoretical background has not yet been provided. The goal of this paper is to provide the background and further improve the performance. First, we show that the fundamental reason contrastive learning methods struggle with long-tailed tasks is that they try to maximize the mutual information maximization between latent features and input data. As ground-truth labels are not considered in the maximization, they are not able to address imbalances between class labels. Rather, we interpret the long-tailed recognition task as a mutual information maximization between latent features and ground-truth labels. This approach integrates contrastive learning and logit adjustment seamlessly to derive a loss function that shows state-of-the-art performance on long-tailed recognition benchmarks. It also demonstrates its efficacy in image segmentation tasks, verifying its versatility beyond image classification.

arXiv:2305.01118

CSP: Self-Supervised Contrastive Spatial Pre-Training for Geospatial-Visual Representations

Geo-tagged images are publicly available in large quantities, whereas labels such as object classes are rather scarce and expensive to collect. Meanwhile, contrastive learning has achieved tremendous success in various natural image and language tasks with limited labeled data. However, existing methods fail to fully leverage geospatial information, which can be paramount to distinguishing objects that are visually similar. To directly leverage the abundant geospatial information associated with images in pre-training, fine-tuning, and inference stages, we present Contrastive Spatial Pre-Training (CSP), a self-supervised learning framework for geo-tagged images. We use a dual-encoder to separately encode the images and their corresponding geo-locations, and use contrastive objectives to learn effective location representations from images, which can be transferred to downstream supervised tasks such as image classification. Experiments show that CSP can improve model performance on both iNat2018 and fMoW datasets. Especially, on iNat2018, CSP significantly boosts the model performance with 10-34% relative improvement with various labeled training data sampling ratios.

arXiv:2305.00002

Galaxy Classification Using Transfer Learning and Ensemble of CNNs With Multiple Colour Spaces

Big data has become the norm in astronomy, making it an ideal domain for computer science research. Astronomers typically classify galaxies based on their morphologies, a practice that dates back to Hubble (1936). With small datasets, classification could be performed by individuals or small teams, but the exponential growth of data from modern telescopes necessitates automated classification methods. In December 2013, Winton Capital, Galaxy Zoo, and the Kaggle team created the Galaxy Challenge, which tasked participants with developing models to classify galaxies. The Kaggle Galaxy Zoo dataset has since been widely used by researchers. This study investigates the impact of colour space transformation on classification accuracy and explores the effect of CNN architecture on this relationship. Multiple colour spaces (RGB, XYZ, LAB, etc.) and CNN architectures (VGG, ResNet, DenseNet, Xception, etc.) are considered, utilizing pre-trained models and weights. However, as most pre-trained models are designed for natural RGB images, we examine their performance with transformed, non-natural astronomical images. We test our hypothesis by evaluating individual networks with RGB and transformed colour spaces and examining various ensemble configurations. A minimal hyperparameter search ensures optimal results. Our findings indicate that using transformed colour spaces in individual networks yields higher validation accuracy, and ensembles of networks and colour spaces further improve accuracy. This research aims to validate the utility of colour space transformation for astronomical image classification and serve as a benchmark for future studies.

arXiv:2305.00104

MMViT: Multiscale Multiview Vision Transformers

We present Multiscale Multiview Vision Transformers (MMViT), which introduces multiscale feature maps and multiview encodings to transformer models. Our model encodes different views of the input signal and builds several channel-resolution feature stages to process the multiple views of the input at different resolutions in parallel. At each scale stage, we use a cross-attention block to fuse information across different views. This enables the MMViT model to acquire complex high-dimensional representations of the input at different resolutions. The proposed model can serve as a backbone model in multiple domains. We demonstrate the effectiveness of MMViT on audio and image classification tasks, achieving state-of-the-art results.

arXiv:2305.00201

Instruction-ViT: Multi-Modal Prompts for Instruction Learning in ViT

Prompts have been proven to play a crucial role in large language models, and in recent years, vision models have also been using prompts to improve scalability for multiple downstream tasks. In this paper, we focus on adapting prompt design based on instruction tuning into a visual transformer model for image classification which we called Instruction-ViT. The key idea is to implement multi-modal prompts (text or image prompt) related to category information to guide the fine-tuning of the model. Based on the experiments of several image captionining tasks, the performance and domain adaptability were improved. Our work provided an innovative strategy to fuse multi-modal prompts with better performance and faster adaptability for visual classification models.

arXiv:2305.00738공식 코드 · Papers with Code

FCA: Taming Long-tailed Federated Medical Image Classification by Classifier Anchoring

Limited training data and severe class imbalance impose significant challenges to developing clinically robust deep learning models. Federated learning (FL) addresses the former by enabling different medical clients to collaboratively train a deep model without sharing data. However, the class imbalance problem persists due to inter-client class distribution variations. To overcome this, we propose federated classifier anchoring (FCA) by adding a personalized classifier at each client to guide and debias the federated model through consistency learning. Additionally, FCA debiases the federated classifier and each client's personalized classifier based on their respective class distributions, thus mitigating divergence. With FCA, the federated feature extractor effectively learns discriminative features suitably globally for federation as well as locally for all participants. In clinical practice, the federated model is expected to be both generalized, performing well across clients, and specialized, benefiting each individual client from collaboration. According to this, we propose a novel evaluation metric to assess models' generalization and specialization performance globally on an aggregated public test set and locally at each client. Through comprehensive comparison and evaluation, FCA outperforms the state-of-the-art methods with large margins for federated long-tailed skin lesion classification and intracranial hemorrhage classification, making it a more feasible solution in clinical settings.

arXiv:2305.00696공식 코드 · Papers with Code

TPMIL: Trainable Prototype Enhanced Multiple Instance Learning for Whole Slide Image Classification

Digital pathology based on whole slide images (WSIs) plays a key role in cancer diagnosis and clinical practice. Due to the high resolution of the WSI and the unavailability of patch-level annotations, WSI classification is usually formulated as a weakly supervised problem, which relies on multiple instance learning (MIL) based on patches of a WSI. In this paper, we aim to learn an optimal patch-level feature space by integrating prototype learning with MIL. To this end, we develop a Trainable Prototype enhanced deep MIL (TPMIL) framework for weakly supervised WSI classification. In contrast to the conventional methods which rely on a certain number of selected patches for feature space refinement, we softly cluster all the instances by allocating them to their corresponding prototypes. Additionally, our method is able to reveal the correlations between different tumor subtypes through distances between corresponding trained prototypes. More importantly, TPMIL also enables to provide a more accurate interpretability based on the distance of the instances from the trained prototypes which serves as an alternative to the conventional attention score-based interpretability. We test our method on two WSI datasets and it achieves a new SOTA. GitHub repository: https://github.com/LitaoYang-Jet/TPMIL

arXiv:2305.00350공식 코드 · Papers with Code

POUF: Prompt-oriented unsupervised fine-tuning for large pre-trained models

Through prompting, large-scale pre-trained models have become more expressive and powerful, gaining significant attention in recent years. Though these big models have zero-shot capabilities, in general, labeled data are still required to adapt them to downstream tasks. To overcome this critical limitation, we propose an unsupervised fine-tuning framework to directly fine-tune the model or prompt on the unlabeled target data. We demonstrate how to apply our method to both language-augmented vision and masked-language models by aligning the discrete distributions extracted from the prompts and target data. To verify our approach's applicability, we conduct extensive experiments on image classification, sentiment analysis, and natural language inference tasks. Across 13 image-related tasks and 15 language-related ones, the proposed approach achieves consistent improvements over the baselines.

arXiv:2304.13726

SamurAI: A Versatile IoT Node With Event-Driven Wake-Up and Embedded ML Acceleration

Increased capabilities such as recognition and self-adaptability are now required from IoT applications. While IoT node power consumption is a major concern for these applications, cloud-based processing is becoming unsustainable due to continuous sensor or image data transmission over the wireless network. Thus optimized ML capabilities and data transfers should be integrated in the IoT node. Moreover, IoT applications are torn between sporadic data-logging and energy-hungry data processing (e.g. image classification). Thus, the versatility of the node is key in addressing this wide diversity of energy and processing needs. This paper presents SamurAI, a versatile IoT node bridging this gap in processing and in energy by leveraging two on-chip sub-systems: a low power, clock-less, event-driven Always-Responsive (AR) part and an energy-efficient On-Demand (OD) part. AR contains a 1.7MOPS event-driven, asynchronous Wake-up Controller (WuC) with a 207ns wake-up time optimized for sporadic computing, while OD combines a deep-sleep RISC-V CPU and 1.3TOPS/W Machine Learning (ML) for more complex tasks up to 36GOPS. This architecture partitioning achieves best in class versatility metrics such as peak performance to idle power ratio. On an applicative classification scenario, it demonstrates system power gains, up to 3.5x compared to cloud-based processing, and thus extended battery lifetime.

논문 ID: 2022.31857

Graph Information Aggregation Cross-Domain Few-Shot Learning for Hyperspectral Image Classification

Most domain adaptation (DA) methods in cross-scene hyperspectral image classification focus on cases where source data (SD) and target data (TD) with the same classes are obtained by the same sensor. However, the classification performance is significantly reduced when there are new classes in TD. In addition, domain alignment, as one of the main approaches in DA, is carried out based on local spatial information, rarely taking into account nonlocal spatial information (nonlocal relationships) with strong correspondence. A graph information aggregation cross-domain few-shot learning (Gia-CFSL) framework is proposed, intending to make up for the above-mentioned shortcomings by combining FSL with domain alignment based on graph information aggregation. SD with all label samples and TD with a few label samples are implemented for FSL episodic training. Meanwhile, intradomain distribution extraction block (IDE-block) and cross-domain similarity aware block (CSA-block) are designed. The IDE-block is used to characterize and aggregate the intradomain nonlocal relationships and the interdomain feature and distribution similarities are captured in the CSA-block. Furthermore, feature-level and distribution-level cross-domain graph alignments are used to mitigate the impact of domain shift on FSL. Experimental results on three public HSI datasets demonstrate the superiority of the proposed method. The codes will be available from the website: https://github.com/YuxiangZhang-BIT/IEEE_TNNLS_Gia-CFSL.

논문 ID: 2019.29463

Deep Metric Learning-Based Feature Embedding for Hyperspectral Image Classification

Learning from a limited number of labeled samples (pixels) remains a key challenge in the hyperspectral image (HSI) classification. To address this issue, we propose a deep metric learning-based feature embedding model, which can meet the tasks both for same- and cross-scene HSI classifications. In the first task, when only a few labeled samples are available, we employ ideas from metric learning based on deep embedding features and make a similarity learning between pairs of samples. In this case, the proposed model can learn well to compare whether two samples belong to the same class. In another task, when an HSI image (target scene) that needs to be classified is not labeled at all, the embedding model can learn from another similar HSI image (source scene) with sufficient labeled samples and then transfer to the target model by using an unsupervised domain adaptation technique, which not only employs the adversarial approach to make the embedding features from the source and target samples indistinguishable but also encourages the target scene's embeddings to form similar clusters with the source scene one. After the domain adaptation between the HSIs of the two scenes is finished, any traditional HSI classifier can be used. In a simple manner, the nearest neighbor (NN) algorithm is selected as the classifier for the classification tasks throughout this article. The experimental results from a series of popular HSIs demonstrate the advantages of the proposed model both in the same- and cross-scene classification tasks.

논문 ID: 2019.29347

HSI-BERT: Hyperspectral Image Classification Using the Bidirectional Encoder Representation From Transformers

Deep learning methods have been widely used in hyperspectral image classification and have achieved state-of-the-art performance. Nonetheless, the existing deep learning methods are restricted by a limited receptive field, inflexibility, and difficult generalization problems in hyperspectral image classification. To solve these problems, we propose HSI-BERT, where BERT stands for bidirectional encoder representations from transformers and HSI stands for hyperspectral imagery. The proposed HSI-BERT has a global receptive field that captures the global dependence among pixels regardless of their spatial distance. HSI-BERT is very flexible and enables the flexible and dynamic input regions. Furthermore, HSI-BERT has good generalization ability because the jointly trained HSI-BERT can be generalized from regions with different shapes without retraining. HSI-BERT is primarily built on a multihead self-attention (MHSA) mechanism in an MHSA layer. Moreover, several attentions are learned by different heads, and each head of the MHSA layer encodes the semantic context-aware representation to obtain discriminative features. Because all head-encoded features are merged, the resulting features exhibit spatial-spectral information that is essential for accurate pixel-level classification. Quantitative and qualitative results demonstrate that HSI-BERT outperforms any other CNN-based model in terms of both classification accuracy and computational time and achieves state-of-the-art performance on three widely used hyperspectral image data sets.

논문 ID: 2013.22756

Feature Extraction of Hyperspectral Images With Image Fusion and Recursive Filtering

Feature extraction is known to be an effective way in both reducing computational complexity and increasing accuracy of hyperspectral image classification. In this paper, a simple yet quite powerful feature extraction method based on image fusion and recursive filtering (IFRF) is proposed. First, the hyperspectral image is partitioned into multiple subsets of adjacent hyperspectral bands. Then, the bands in each subset are fused together by averaging, which is one of the simplest image fusion methods. Finally, the fused bands are processed with transform domain recursive filtering to get the resulting features for classification. Experiments are performed on different hyperspectral images, with the support vector machines (SVMs) serving as the classifier. By using the proposed method, the accuracy of the SVM classifier can be improved significantly. Furthermore, compared with other hyperspectral classification methods, the proposed IFRF method shows outstanding performance in terms of classification accuracy and computational efficiency.

arXiv:2304.13775

Advancing Ischemic Stroke Diagnosis: A Novel Two-Stage Approach for Blood Clot Origin Identification

An innovative two-stage methodology for categorizing blood clot origins is presented in this paper, which is important for the diagnosis and treatment of ischemic stroke. First, a background classifier based on MobileNetV3 segments big whole-slide digital pathology images into numerous tiles to detect the presence of cellular material. After that, different pre-trained image classification algorithms are fine-tuned to determine the origin of blood clots. Due to complex blood flow dynamics and limitations in conventional imaging methods such as computed tomography (CT), magnetic resonance imaging (MRI), and ultrasound, identifying the sources of blood clots is a challenging task. Although these techniques are useful for identifying blood clots, they are not very good at determining how they originated. To address these challenges, our method makes use of robust computer vision models that have been refined using information from whole-slide digital pathology images. Out of all the models tested, the PoolFormer \cite{yu2022metaformer} performs better than the others, with 93.4\% accuracy, 93.4\% precision, 93.4\% recall, and 93.4\% F1-score. Moreover, it achieves the good weighted multi-class logarithmic loss (WMCLL) of 0.4361, which emphasizes how effective it is in this particular application. These encouraging findings suggest that our approach can successfully identify the origin of blood clots in a variety of vascular locations, potentially advancing ischemic stroke diagnosis and treatment approaches.

arXiv:2304.13538

ElegansNet: a brief scientific report and initial experiments

This research report introduces ElegansNet, a neural network that mimics real-world neuronal network circuitry, with the goal of better understanding the interplay between connectome topology and deep learning systems. The proposed approach utilizes the powerful representational capabilities of living beings' neuronal circuitry to design and generate improved deep learning systems with a topology similar to natural networks. The Caenorhabditis elegans connectome is used as a reference due to its completeness, reasonable size, and functional neuron classes annotations. It is demonstrated that the connectome of simple organisms exhibits specific functional relationships between neurons, and once transformed into learnable tensor networks and integrated into modern architectures, it offers bio-plausible structures that efficiently solve complex tasks. The performance of the models is demonstrated against randomly wired networks and compared to artificial networks ranked on global benchmarks. In the first case, ElegansNet outperforms randomly wired networks. Interestingly, ElegansNet models show slightly similar performance with only those based on the Watts-Strogatz small-world property. When compared to state-of-the-art artificial neural networks, such as transformers or attention-based autoencoders, ElegansNet outperforms well-known deep learning and traditional models in both supervised image classification tasks and unsupervised hand-written digits reconstruction, achieving top-1 accuracy of 99.99% on Cifar10 and 99.84% on MNIST Unsup on the validation sets.

arXiv:2304.13181

Sample-Specific Debiasing for Better Image-Text Models

Self-supervised representation learning on image-text data facilitates crucial medical applications, such as image classification, visual grounding, and cross-modal retrieval. One common approach involves contrasting semantically similar (positive) and dissimilar (negative) pairs of data points. Drawing negative samples uniformly from the training data set introduces false negatives, i.e., samples that are treated as dissimilar but belong to the same class. In healthcare data, the underlying class distribution is nonuniform, implying that false negatives occur at a highly variable rate. To improve the quality of learned representations, we develop a novel approach that corrects for false negatives. Our method can be viewed as a variant of debiased contrastive learning that uses estimated sample-specific class probabilities. We provide theoretical analysis of the objective function and demonstrate the proposed approach on both image and paired image-text data sets. Our experiments illustrate empirical advantages of sample-specific debiasing.