Fine-grained Datasets Research Articles

A finer-grained high altitude EEG dataset for hypoxia levels assessment

The study reports on a high-altitude EEG dataset comprising 64-channel EEG signals from 23 subjects, aiming at achieving a finer-grained assessment of hypoxia levels. Four hypoxia levels were induced by creating a gradient of oxygen partial pressure through changes in altitude and external hypoxia stimulation. The dataset was collected in a hypoxic chamber that simulates altitude changes, allowing for a refined classification of different hypoxia levels based on ranges of oxygen saturation. The total recorded EEG data amounts to approximately 10.25 hours. Validation results indicate that the four hypoxia levels can be effectively recognized using EEG signals. Compared to binary classification, our fine-grained dataset allows for more precise detection of hypoxia levels. This dataset is anticipated to have significant research and practical value in developing accurate methods for identifying hypoxia levels. As a valuable and standardized resource, it will enable extensive analysis and comparison for researchers in the field of high-altitude hypoxia.

One size does not fit all in evaluating model selection scores for image classification

Selecting pretrained models for image classification often involves computationally intensive finetuning. This study addresses a research gap in the standardized evaluation of transferability scores, which could simplify model selection by ranking pretrained models without exhaustive finetuning. The motivation is to reduce the computational burden of model selection through a consistent approach that guides practitioners in balancing accuracy and efficiency across tasks. This study evaluates 14 transferability scores on 11 benchmark datasets. It includes both Convolutional Neural Network (CNN) and Vision Transformer (ViT) models and ensures consistency in experimental conditions to counter the variability in previous research. Key findings reveal significant variability in score effectiveness based on dataset characteristics (e.g., fine-grained versus coarse-grained classes) and model architectures. ViT models generally show superior transferability, especially for fine-grained datasets. While no single score is best in all cases, some scores excel in specific contexts. In addition to predictive accuracy, the study also evaluates computational efficiency and identifies scores that are suitable for resource-constrained scenarios. This research provides insights for selecting appropriate transferability scores to optimize model selection strategies to facilitate efficient deployment in practice.

Housing Market Segmentation as a Driver of Urban Micro-Segregation? An In-Depth Analysis of Two Viennese Districts

The concept of segregation analyses the unequal distribution of social groups between neighbourhoods. It rests on two assumptions: that of homogeneous neighbourhoods and of a market liberal housing system. Both assumptions are applicable the context of American cities, but they display severe limitations when applied to the European context. Vienna’s housing market is particularly highly segmented, not only throughout the city as a whole but also within neighbourhoods. In the densely built-up area, residential buildings of different segments with different underlying rent regulations and entry barriers can be found side by side. Therefore, buildings are expected to show varying tenant and owner structures, which undermines the idea of a homogeneous neighbourhood. Against this background, we analyse at the micro scale small neighbourhoods defined by 100 m grid cells in a case study of two inner-city Viennese districts (districts 6 and 7) characterised by a particularly vivid housing-transformation and commodification dynamic. Using a novel and fine-grained dataset combining building information with the socio-economic data of households, we investigate the patterns and dynamics of income inequality and income segregation, as well as the relationship between housing market segments and socio-economic patterns. As data comprise two cross-sections for the years 2011 and 2020/21, changes in the neighbourhoods during the house-price boom period are also considered. This leads us to ask the question: How do housing market segmentation and its related changes affect income inequality and segregation at the micro scale? Our analysis delivers two main results: Firstly, we show the existence of marked social variation and related dynamics at the micro scale, even within a small urban area. Secondly, we show that the spatial distribution of housing market segments has a strong impact on income inequality in the neighbourhood.

Analyzing Mobility Patterns at Scale in Pandemic Scenarios Leveraging the Mobile Network Ecosystem

The ubiquity and pervasiveness of mobile network technologies has made them so deeply ingrained in our everyday lives that by interacting with them for very simple purposes (e.g., messaging or browsing the Internet), we produce an unprecedented amount of data that can be analyzed to understand our behavior. While this practice has been extensively adopted by telcos and big tech companies in the last few years, this condition, which was unimaginable just 20 years ago, has only been mildly exploited to fight the COVID-19 pandemic. In this paper, we discuss the possible alternatives that we could leverage in the current mobile network ecosystem to provide regulators and epidemiologists with the right understanding of our mobility patterns, to maximize the efficiency and extent of the introduced countermeasures. To validate our analysis, we dissect a fine-grained dataset of user positions in two major European countries severely hit by the pandemic. The potential of using these data, harvested employing traditional mobile network technologies, is unveiled through two exemplary cases that tackled macro and microscopic aspects.

Feature extractor optimization for discriminative representations in Generalized Category Discovery

Generalized Category Discovery (GCD) task involves transferring knowledge from labeled known categories to recognize both known and novel categories within an unlabeled dataset. A significant challenge arises from the lack of prior information for novel categories. To address this, we develop a feature extractor that can learn discriminative features for both known and novel categories. Our approach leverages the observation that similar samples often belong to the same class. We construct a similarity matrix and employ similarity contrastive loss to increase the similarity between similar samples in the feature space. Additionally, we incorporate cluster labels to further refine the feature extractor, utilizing K-means clustering to assign these labels to unlabeled data, providing valuable supervision. Our feature extractor is optimized through the utilization of instance-level contrastive learning and class-level contrastive learning constraints. These constraints promote similarity maximization in both the instance space and the label space for instances sharing the same pseudo-labels. These three components complement each other, facilitating the learning of discriminative representations for both known and novel categories. Through comprehensive evaluations of generic image recognition datasets and challenging fine-grained datasets, we demonstrate that our proposed method achieves state-of-the-art performance.

TST_MFL: Two-stage training based metric fusion learning for few-shot image classification

Addressing the limitations of most few-shot learning (FSL) methods, particularly their insufficient single-feature discriminability and generalization during pre-training encoding, this paper introduces a novel approach: Two-Stage Training based Metric Fusion Learning (TST_MFL) for few-shot image classification. TST_MFL innovatively integrates a two-stage fusion metric learning strategy, enhancing both pre-training and meta-training phases to overcome the prevalent focus on single-level feature extraction in current few-shot image classification methods. In pre-training stage, we employ prediction distillation loss and local difference loss to facilitate mutual learning between global and local features, thus obtaining more discriminative features. Moreover, meta-training further enhances feature generalization and integrates global–local metrics to improve few-shot classification accuracy. Our approach demonstrates competitive performance across three fine-grained datasets and one coarse-grained dataset. Notably, experimental results in 1-shot and 5-shot settings of fine-grained datasets underscore the superiority of TST_MFL. The code for this work is available at: https://github.com/ZitZhengWang/TST_MFL.

MICCF: A Mutual Information Constrained Clustering Framework for Learning Clustering-Oriented Feature Representations

Deep clustering is a crucial task in machine learning and data mining that focuses on acquiring feature representations conducive to clustering. Previous research relies on self-supervised representation learning for general feature representations, such features may not be optimally suited for downstream clustering tasks. In this paper, we introduce MICCF, a framework designed to bridge this gap and enhance clustering performance. MICCF enhances feature representations by combining mutual information at different levels and employs an auxiliary alignment mutual information module for learning clustering-oriented features. To be specific, we propose a dual mutual information constraints module, incorporating minimal mutual information constraints at the feature level and maximal mutual information constraints at the instance level. This reduction in feature redundancy encourages the neural network to extract more discriminative features, while maximization ensures more unbiased and robust representations. To obtain clustering-oriented representations, the auxiliary alignment mutual information module utilizes pseudo-labels to maximize mutual information through a multi-classifier network, aligning features with the clustering task. The main network and the auxiliary one work in synergy to jointly optimize feature representations that are well-suited for the clustering task. We validate the effectiveness of our method through extensive experiments on six benchmark datasets. The results indicate that our method performs well in most scenarios, particularly on fine-grained datasets, where our approach effectively distinguishes subtle differences between closely related categories. Notably, our approach achieved a remarkable accuracy of 96.4% on the ImageNet-10 dataset, surpassing other comparison methods. The code is available at https://github.com/Li-Hyn/MICCF.git .

A fine-grained dataset for sewage outfalls objective detection in natural environments

Pollution sources release contaminants into water bodies via sewage outfalls (SOs). Using high-resolution images to interpret SOs is laborious and expensive because it needs specific knowledge and must be done by hand. Integrating unmanned aerial vehicles (UAVs) and deep learning technology could assist in constructing an automated effluent SOs detection tool by gaining specialized knowledge. Achieving this objective requires high-quality image datasets for model training and testing. However, there is no satisfactory dataset of SOs. This study presents a high-quality dataset named the images for sewage outfalls objective detection (iSOOD). The 10481 images in iSOOD were captured using UAVs and handheld cameras by individuals from the river basin in China. This study has carefully annotated these images to ensure accuracy and consistency. The iSOOD has undergone technical validation utilizing the YOLOv10 series objective detection model. Our study could provide high-quality SOs datasets for enhancing deep-learning models with UAVs to achieve efficient and intelligent river basin management.

Multi-FusNet: fusion mapping of features for fine-grained image retrieval networks.

As the diversity and volume of images continue to grow, the demand for efficient fine-grained image retrieval has surged across numerous fields. However, the current deep learning-based approaches to fine-grained image retrieval often concentrate solely on the top-layer features, neglecting the relevant information carried in the middle layer, even though these information contains more fine-grained identification content. Moreover, these methods typically employ a uniform weighting strategy during hash code mapping, risking the loss of critical region mapping-an irreversible detriment to fine-grained retrieval tasks. To address the above problems, we propose a novel method for fine-grained image retrieval that leverage feature fusion and hash mapping techniques. Our approach harnesses a multi-level feature cascade, emphasizing not just top-layer but also intermediate-layer image features, and integrates a feature fusion module at each level to enhance the extraction of discriminative information. In addition, we introduce an agent self-attention architecture, marking its first application in this context, which steers the model to prioritize on long-range features, further avoiding the loss of critical regions of the mapping. Finally, our proposed model significantly outperforms existing state-of-the-art, improving the retrieval accuracy by an average of 40% for the 12-bit dataset, 22% for the 24-bit dataset, 16% for the 32-bit dataset, and 11% for the 48-bit dataset across five publicly available fine-grained datasets. We also validate the generalization ability and performance stability of our proposed method by another five datasets and statistical significance tests. Our code can be downloaded from https://github.com/BJFU-CS2012/MuiltNet.git.

Pseudo-set Frequency Refinement architecture for fine-grained few-shot class-incremental learning

Few-shot class-incremental learning was introduced to solve the model adaptation problem for new incremental classes with only a few examples while still remaining effective for old data. Although recent state-of-the-art methods make some progress in improving system robustness on common datasets, they fail to work on fine-grained datasets where inter-class differences are small. The problem is mainly caused by: (1) the overlapping of new data and old data in the feature space during incremental learning, which means old samples can be falsely classified as newly introduced classes and induce catastrophic forgetting phenomena; (2) lacking discriminative feature learning ability to identify fine-grained objects. In this paper, a novel Pseudo-set Frequency Refinement (PFR) architecture is proposed to tackle these problems. We design a pseudo-set training strategy to mimic the incremental learning scenarios so that the model can better adapt to novel data in future incremental sessions. Furthermore, separate adaptation tasks are developed by utilizing frequency-based information to refine the original features and address the above challenging problems. More specifically, the high and low-frequency components of the images are employed to enrich the discriminative feature analysis ability and incremental learning ability of the model respectively. The refined features are used to perform inter-class and inter-set analyses. Extensive experiments show that the proposed method consistently outperforms the state-of-the-art methods on four fine-grained datasets.

Distributed training of CosPlace for large-scale visual place recognition.

Visual place recognition (VPR) is a popular computer vision task aimed at recognizing the geographic location of a visual query, usually within a tolerance of a few meters. Modern approaches address VPR from an image retrieval standpoint using a kNN on top of embeddings extracted by a deep neural network from both the query and images in a database. Although most of these approaches rely on contrastive learning, which limits their ability to be trained on large-scale datasets (due to mining), the recently reported CosPlace proposes an alternative training paradigm using a classification task as the proxy. This has been shown to be effective in expanding the potential of VPR models to learn from large-scale and fine-grained datasets. In this work, we experimentally analyze CosPlace from a continual learning perspective and show that its sequential training procedure leads to suboptimal results. As a solution, we propose a different formulation that not only solves the pitfalls of the original training strategy effectively but also enables faster and more efficient distributed training. Finally, we discuss the open challenges in further speeding up large-scale image retrieval for VPR.

A Comparative Analysis of Random Forest, XGBoost, and LightGBM Algorithms for Emotion Classification in Reddit Comments

This research aims to compare the performance of three classification algorithms, namely Random Forest, XGBoost, and LightGBM, in classifying emotions in Reddit comments. Emotion classification in Reddit comments is a complex classification problem due to its numerous variations and ambiguities. This research utilizes the GoEmotions Fine-Grained dataset, filtered down to 7,325 Reddit comments with 5 different basic emotion labels. In this study, data preprocessing steps, feature extraction using CountVectorizer and TF-IDF, and hyperparameter tuning using GridSearchCV for each algorithm are conducted. Subsequently, model evaluation is performed using Cross-Validation and confusion matrix. The results of the study indicate that Random Forest outperforms the XGBoost and LightGBM algorithm with an accuracy of 75.38% compared to XGBoost with 69.05% accuracy and LightGBM with 66.63% accuracy.

AlpaPICO: Extraction of PICO frames from clinical trial documents using LLMs

In recent years, there has been a surge in the publication of clinical trial reports, making it challenging to conduct systematic reviews. Automatically extracting Population, Intervention, Comparator, and Outcome (PICO) from clinical trial studies can alleviate the traditionally time-consuming process of manually scrutinizing systematic reviews. Existing approaches of PICO frame extraction involves supervised approach that relies on the existence of manually annotated data points in the form of BIO label tagging. Recent approaches, such as In-Context Learning (ICL), which has been shown to be effective for a number of downstream NLP tasks, require the use of labeled examples. In this work, we adopt ICL strategy by employing the pretrained knowledge of Large Language Models (LLMs), gathered during the pretraining phase of an LLM, to automatically extract the PICO-related terminologies from clinical trial documents in unsupervised set up to bypass the availability of large number of annotated data instances. Additionally, to showcase the highest effectiveness of LLM in oracle scenario where large number of annotated samples are available, we adopt the instruction tuning strategy by employing Low Rank Adaptation (LORA) to conduct the training of gigantic model in low resource environment for the PICO frame extraction task. More specifically, both of the proposed frameworks utilize AlpaCare as base LLM which employs both few-shot in-context learning and instruction tuning techniques to extract PICO-related terms from the clinical trial reports. We applied these approaches to the widely used coarse-grained datasets such as EBM-NLP, EBM-COMET and fine-grained datasets such as EBM-NLPrev and EBM-NLPh. Our empirical results show that our proposed ICL-based framework produces comparable results on all the version of EBM-NLP datasets and the proposed instruction tuned version of our framework produces state-of-the-art results on all the different EBM-NLP datasets. Our project is available at https://github.com/shrimonmuke0202/AlpaPICO.git.

Self-reconstruction network for fine-grained few-shot classification

Metric-based methods are one of the most common methods to solve the problem of few-shot image classification. However, traditional metric-based few-shot methods suffer from overfitting and local feature misalignment. The recently proposed feature reconstruction-based approach, which reconstructs query image features from the support set features of a given class and compares the distance between the original query features and the reconstructed query features as the classification criterion, effectively solves the feature misalignment problem. However, the issue of overfitting still has not been considered. To this end, we propose a self-reconstruction metric module for diversifying query features and a restrained cross-entropy loss for avoiding over-confident predictions. By introducing them, the proposed self-reconstruction network can effectively alleviate overfitting. Extensive experiments on five benchmark fine-grained datasets demonstrate that our proposed method achieves state-of-the-art performance on both 5-way 1-shot and 5-way 5-shot classification tasks. Code is available at https://github.com/liz-lut/SRM-main.

Generalizability and robustness evaluation of attribute-based zero-shot learning

In the field of deep learning, large quantities of data are typically required to effectively train models. This challenge has given rise to techniques like zero-shot learning (ZSL), which trains models on a set of “seen” classes and evaluates them on a set of “unseen” classes. Although ZSL has shown considerable potential, particularly with the employment of generative methods, its generalizability to real-world scenarios remains uncertain.The hypothesis of this work is that the performance of ZSL models is systematically influenced by the chosen “splits”; in particular, the statistical properties of the classes and attributes used in training. In this paper, we test this hypothesis by introducing the concepts of generalizability and robustness in attribute-based ZSL and carry out a variety of experiments to stress-test ZSL models against different splits. Our aim is to lay the groundwork for future research on ZSL models’ generalizability, robustness, and practical applications.We evaluate the accuracy of state-of-the-art models on benchmark datasets and identify consistent trends in generalizability and robustness. We analyze how these properties vary based on the dataset type, differentiating between coarse- and fine-grained datasets, and our findings indicate significant room for improvement in both generalizability and robustness. Furthermore, our results demonstrate the effectiveness of dimensionality reduction techniques in improving the performance of state-of-the-art models in fine-grained datasets.

Concept-Guided Prompt Learning for Generalization in Vision-Language Models

Contrastive Language-Image Pretraining (CLIP) model has exhibited remarkable efficacy in establishing cross-modal connections between texts and images, yielding impressive performance across a broad spectrum of downstream applications through fine-tuning. However, for generalization tasks, the current fine-tuning methods for CLIP, such as CoOp and CoCoOp, demonstrate relatively low performance on some fine-grained datasets. We recognize the underlying reason is that these previous methods only projected global features into the prompt, neglecting the various visual concepts, such as colors, shapes, and sizes, which are naturally transferable across domains and play a crucial role in generalization tasks. To address this issue, in this work, we propose Concept-Guided Prompt Learning (CPL) for vision-language models. Specifically, we leverage the well-learned knowledge of CLIP to create a visual concept cache to enable conceptguided prompting. In order to refine the text features, we further develop a projector that transforms multi-level visual features into text features. We observe that this concept-guided prompt learning approach is able to achieve enhanced consistency between visual and linguistic modalities. Extensive experimental results demonstrate that our CPL method significantly improves generalization capabilities compared to the current state-of-the-art methods.

From Coarse to Fine: A Distillation Method for Fine-Grained Emotion-Causal Span Pair Extraction in Conversation

We study the problem of extracting emotions and the causes behind these emotions in conversations. Existing methods either tackle them separately or jointly model them at the coarse-grained level of emotions (fewer emotion categories) and causes (utterance-level causes). In this work, we aim to jointly extract more fine-grained emotions and causes. We construct a fine-grained dataset FG-RECCON, includes 16 fine-grained emotion categories and span-level causes. To further improve the fine-grained extraction performance, we propose to utilize the casual discourse knowledge in a knowledge distillation way. Specifically, the teacher model learns to predict causal connective words between utterances, and then guides the student model in identifying both the fine-grained emotion labels and causal spans. Experimental results demonstrate that our distillation method achieves state-of-the-art performance on both RECCON and FG-RECCON dataset.

Tolerant Self-Distillation for image classification

Deep neural networks tend to suffer from the overfitting issue when the training data are not enough. In this paper, we introduce two metrics from the intra-class distribution of correct-predicted and incorrect-predicted samples to provide a new perspective on the overfitting issue. Based on it, we propose a knowledge distillation approach without pretraining a teacher model in advance named Tolerant Self-Distillation (TSD) for alleviating the overfitting issue. It introduces an online updating memory and selectively stores the class predictions of the samples from the past iterations, making it possible to distill knowledge across the iterations. Specifically, the class predictions stored in the memory bank serve as the soft labels for supervising the samples from the same class for the current iteration in a reverse way, i.e. the correct-predicted samples are supervised with the incorrect predictions while the incorrect-predicted samples are supervised with the correct predictions. Consequently, the premature convergence issue caused by the over-confident samples would be mitigated, which helps the model to converge to a better local optimum. Extensive experimental results on several image classification benchmarks, including small-scale, large-scale, and fine-grained datasets, demonstrate the superiority of the proposed TSD.

Multidimensional Refinement Graph Convolutional Network With Robust Decouple Loss for Fine-Grained Skeleton-Based Action Recognition.

Graph convolutional networks (GCNs) have been widely used in skeleton-based action recognition. However, existing approaches are limited in fine-grained action recognition due to the similarity of interclass data. Moreover, the noisy data from pose extraction increase the challenge of fine-grained recognition. In this work, we propose a flexible attention block called channel-variable spatial-temporal attention (CVSTA) to enhance the discriminative power of spatial-temporal joints and obtain a more compact intraclass feature distribution. Based on CVSTA, we construct a multidimensional refinement GCN (MDR-GCN) that can improve the discrimination among channel-, joint-, and frame-level features for fine-grained actions. Furthermore, we propose a robust decouple loss (RDL) that significantly boosts the effect of the CVSTA and reduces the impact of noise. The proposed method combining MDR-GCN with RDL outperforms the known state-of-the-art skeleton-based approaches on fine-grained datasets, FineGym99 and FSD-10, and also on the coarse NTU-RGB + D 120 dataset and NTU-RGB + D X-view version. Our code is publicly available at https://github.com/dingyn-Reno/MDR-GCN.

Panoptic Perception: A Novel Task and Fine-grained Dataset for Universal Remote Sensing Image Interpretation

Panoptic Perception: A Novel Task and Fine-grained Dataset for Universal Remote Sensing Image Interpretation