Query Features Research Articles

GRASSIUS 2.0: A gene regulatory information knowledgebase for maize and other grasses

Grass species, which include the major cereal crops maize, wheat, rice, and sugarcane, are an integral part of our global agriculture and source of food and energy for a growing world population. GRASSIUS was established as a knowledgebase for transcription factors (TFs) and coregulators (CoRegs) in maize and several other species in the grass family. TFs are a primary component of the gene regulatory networks (GRNs) and the underlying gene regulatory grids (GRGs) that govern all aspects of plant growth and metabolism. GRASSIUS is the source for all information pertaining to the maize TFome collection, which serves as a powerful resource for the discovery of GRNs in maize and other cereals [1,2]. Here we describe the release of the GRASSIUS 2.0 knowledgebase (www.grassius.org) with updated data, query, and tool features, as well as the ability to expand to accommodate future datasets. The membership and annotation of all TF and CoReg families has been updated and revised to include gene models from v3, v4, and v5, of the maize B73 genome and recent genome versions of rice, sorghum, Brachypodium, and sugarcane. A translation tool enables cross referencing of Gene IDs between versions of the maize genome. Protein-DNA interactions (PDIs) have been added incorporating results derived from various gene- and TF-centered PDI discovery tools and visualized through a new web-based interface. A filtering tools permits the selection and visualization of PDIs within a ± 2 kb distance from the transcriptional start site (TSS) of a gene of interest. A new BLAST (Basic Local Alignment Search Tool) tool facilitates searching of the maize TFome as well as v3, v4, and v5 TF gene model sets. Lastly, we describe the methodology used to implement GRASSIUS 2.0 which can guide others in developing and updating similar plant gene regulatory knowledgebases.

Query-aware multi-scale proposal network for weakly supervised temporal sentence grounding in videos

Recently, weakly supervised temporal sentence grounding in videos (TSGV) has attracted extensive attention because it does not require precise start-end time annotations during training, and it can quickly retrieve interesting segments according to user needs. In weakly supervised TSGV, query reconstruction (QR)-based methods are the current mainstream, and the quality of proposals determines their performance. QR-based methods have two problems in proposal quality. First, a multi-modal global token is usually mapped to proposals with limited duration diversity, making it difficult to capture relevant segments at varying durations in real scenarios. Additionally, Gaussian functions are typically used to generate relatively fixed weights for frames within proposals, which weigh the original video features to generate proposal-specific features. This results in query-irrelevant frames affecting the discrimination of the proposal features. In this study, we propose a query-aware multi-scale proposal network (QMN). Initially, pre-trained encoders are used to extract video and query features. Subsequently, a multi-scale proposal generation module is designed to refine video features guided by queries and diversify the duration of the proposal. This module performs multi-modal interaction and multi-scale modeling to obtain proposals of different durations. Furthermore, to extract discriminative proposal features and enhance the modeling of proposal frame correlation, a query-aware weight generator is constructed to learn frame weights to suppress query-irrelevant frame representations through contrastive learning. Finally, the masked query is reconstructed using the proposal features to select the best proposal. The effectiveness of the proposed QMN is verified through experiments on the Charades-STA and ActivityNet-Captions datasets.

Found missing semantics: Supplemental prototype network for few-shot semantic segmentation

Few-shot semantic segmentation alleviates the problem of massive data requirements and high costs in semantic segmentation tasks. By learning from support set, few-shot semantic segmentation can segment new classes. However, existing few-shot semantic segmentation methods suffer from information loss during the process of mask average pooling. To address this problem, we propose a supplemental prototype network (SPNet). The SPNet aggregates the lost information from global prototypes to create a supplemental prototype, which enhances the segmentation performance for the current class. In addition, we utilize mutual attention to enhance the similarity between the support and the query feature maps, allowing the model to better identify the target to be segmented. Finally, we introduce a Self-correcting auxiliary, which utilizes the data more effectively to improve segmentation accuracy. We conducted extensive experiments on PASCAL-5i and COCO-20i, which demonstrated the effectiveness of SPNet. And our method achieved state-of-the-art results in the 1-shot and 5-shot semantic segmentation settings.

Using query semantic and feature transfer fusion to enhance cardinality estimating of property graph queries

With the increasing complexity and diversity of query tasks, cardinality estimation has become one of the most challenging problems in query optimization. In this study, we propose an efficient and accurate cardinality estimation method to address the cardinality estimation problem in property graph queries, particularly in response to the current research gap regarding the neglect of contextual semantic features. We first propose formal representations of the property graph query and define its cardinality estimation problem. Then, through the query featurization, we transform the query into a vector representation that can be learned by the estimation model, and enrich the feature vector representation by the context semantic information of the query. We finally propose an estimation model for property graph queries, specifically introducing a feature information transfer module to dynamically control the information flow meanwhile achieving the model’s feature fusion and inference. Experimental results on three datasets show that the estimation model can accurately and efficiently estimate the cardinality of property graph queries, the mean Q_error and RMSE are reduced by about 30% and 25% than the state-of-art estimation models. The context semantics features of queries can improve the model’s estimation accuracy, the mean Q_error result is reduced by about 20% and the RMSE result is about 5%.

On filling the intra-class and inter-class gaps for few-shot segmentation

Few-shot segmentation (FSS) has made remarkable success through prototypical learning. However, owing to the scarcity of support data, FSS methods continue to suffer from huge intra-class and inter-class gaps. In this paper, we introduce a unified network, termed FGNet++, aimed at addressing both gaps. FGNet++ comprises a Self-Adaptive Module (SAM) to emphasize the query features and generate enhanced prototypes for self-alignment. These prototypes contain the intrinsic information of each query sample, thereby mitigating the intra-class appearance gaps. Moreover, we augment SAM with a Feature Transformation Module (FTM) to further reduce intra-class appearance discrepancies. On the other hand, we introduce an Inter-class Feature Separation Module (IFSM) dedicated to bridge the inter-class gap. With a B-SLIC based background loss and a cross-category loss for training, IFSM makes it easy to distinguish the feature space of the target class from other non-target classes. Furthermore, to assess the generality of our approach, we extend our work to 3D point cloud few-shot segmentation and present FGNet-3D. Experimental results demonstrate that our method successfully mitigates both intra-class and inter-class gaps in FSS through SAM and IFSM, respectively, and achieves state-of-the-art performances on multiple datasets compared with previous approaches.

Layer-Wise Mutual Information Meta-Learning Network for Few-Shot Segmentation.

The goal of few-shot segmentation (FSS) is to segment unlabeled images belonging to previously unseen classes using only a limited number of labeled images. The main objective is to transfer label information effectively from support images to query images. In this study, we introduce a novel meta-learning framework called layer-wise mutual information (LayerMI), which enhances the propagation of label information by maximizing the mutual information (MI) between support and query features at each layer. Our approach involves the utilization of a LayerMI Block based on information-theoretic co-clustering. This block performs online co-clustering on the joint probability distribution obtained from each layer, generating a target-specific attention map. The LayerMI Block can be seamlessly integrated into the meta-learning framework and applied to all convolutional neural network (CNN) layers without altering the training objectives. Notably, the LayerMI Block not only maximizes MI between support and query features but also facilitates internal clustering within the image. Extensive experiments demonstrate that LayerMI significantly enhances the performance of baseline and achieves competitive performance compared to state-of-the-art methods on three challenging benchmarks: PASCAL-5 i , COCO-20 i , and FSS-1000.

HQ-Net: A heatmap-based query backbone for point cloud understanding

With the advancement of sensor technology, including LiDAR and RGB-D cameras, tasks related to point cloud understanding have gained widespread application. The sampling and learning of feature points play a crucial role in point cloud understanding, representing a fundamental and essential aspect of 3D computer vision. In the realm of point cloud processing, traditional methods such as Farthest Point Sampling (FPS) are frequently employed to select feature points. These points are then used in conjunction with effective feature aggregation operators to facilitate reasoning. However, these traditional sampling methods primarily focus on the Euclidean distance between points, lacking a comprehensive consideration of point cloud features. To address this limitation, we introduce HQ-Net, a novel point cloud understanding model designed to optimize feature point sampling using a Heatmap-based Query Backbone. The Heatmap-based Query Backbone utilizes the Heatmap Module to select category-specific point features, effectively generating a heatmap tailored to the category features and automatically extracting features from the heatmap. Subsequently, the proposed Query Encoder is employed to facilitate the interaction between point features and query features, providing mutual guidance and enhancing the model’s ability to capture spatial structure and semantic information within point clouds. Additionally, we propose task-specific heads, such as classification and segmentation heads for different tasks, which can be seamlessly integrated with the backbone to achieve excellent performance. In various point cloud understanding tasks, including shape classification, part segmentation, and semantic segmentation, HQ-Net demonstrates excellent performance.

A mutual reconstruction network model for few-shot classification of histological images: addressing interclass similarity and intraclass diversity.

The automated classification of histological images is crucial for the diagnosis of cancer. The limited availability of well-annotated datasets, especially for rare cancers, poses a significant challenge for deep learning methods due to the small number of relevant images. This has led to the development of few-shot learning approaches, which bear considerable clinical importance, as they are designed to overcome the challenges of data scarcity in deep learning for histological image classification. Traditional methods often ignore the challenges of intraclass diversity and interclass similarities in histological images. To address this, we propose a novel mutual reconstruction network model, aimed at meeting these challenges and improving the few-shot classification performance of histological images. The key to our approach is the extraction of subtle and discriminative features. We introduce a feature enhancement module (FEM) and a mutual reconstruction module to increase differences between classes while reducing variance within classes. First, we extract features of support and query images using a feature extractor. These features are then processed by the FEM, which uses a self-attention mechanism for self-reconstruction of features, enhancing the learning of detailed features. These enhanced features are then input into the mutual reconstruction module. This module uses enhanced support features to reconstruct enhanced query features and vice versa. The classification of query samples is based on weighted calculations of the distances between query features and reconstructed query features and between support features and reconstructed support features. We extensively evaluated our model using a specially created few-shot histological image dataset. The results showed that in a 5-way 10-shot setup, our model achieved an impressive accuracy of 92.09%. This is a 23.59% improvement in accuracy compared to the model-agnostic meta-learning (MAML) method, which does not focus on fine-grained attributes. In the more challenging, 5-way 1-shot setting, our model also performed well, demonstrating a 18.52% improvement over the ProtoNet, which does not address this challenge. Additional ablation studies indicated the effectiveness and complementary nature of each module and confirmed our method's ability to parse small differences between classes and large variations within classes in histological images. These findings strongly support the superiority of our proposed method in the few-shot classification of histological images. The mutual reconstruction network provides outstanding performance in the few-shot classification of histological images, successfully overcoming the challenges of similarities between classes and diversity within classes. This marks a significant advancement in the automated classification of histological images.

Bilaterally Normalized Scale-Consistent Sinkhorn Distance for Few-Shot Image Classification.

Few-shot image classification aims at exploring transferable features from base classes to recognize images of the unseen novel classes with only a few labeled images. Existing methods usually compare the support features and query features, which are implemented by either matching the global feature vectors or matching the local feature maps at the same position. However, few labeled images fail to capture all the diverse context and intraclass variations, leading to mismatch issues for existing methods. On one hand, due to the misaligned position and cluttered background, existing methods suffer from the object mismatch issue. On the other hand, due to the scale inconsistency between images, existing methods suffer from the scale mismatch issue. In this article, we propose the bilaterally normalized scale-consistent Sinkhorn distance (BSSD) to solve these issues. First, instead of same-position matching, we use the Sinkhorn distance to find an optimal matching between images, mitigating the object mismatch caused by misaligned position. Meanwhile, we propose the intraimage and interimage attentions as the bilateral normalization on the Sinkhorn distance to suppress the object mismatch caused by background clutter. Second, local feature maps are enhanced with the multiscale pooling strategy, making the Sinkhorn distance possible to find a consistent matching scale between images. Experimental results show the effectiveness of the proposed approach, and we achieve the state-of-the-art on three few-shot benchmarks.

Prototype as query for few shot semantic segmentation

Few-shot Semantic Segmentation (FSS) was proposed to segment unseen classes in a query image, referring to only a few annotated examples named support images. One of the characteristics of FSS is spatial inconsistency between query and support targets, e.g., texture or appearance. This greatly challenges the generalization ability of methods for FSS, which requires to effectively exploit the dependency of the query image and the support examples. Most existing methods abstracted support features into prototype vectors and implemented the interaction with query features using cosine similarity or feature concatenation. However, this simple interaction may not capture spatial details in query features. To address this limitation, some methods utilized pixel-level support information by computing pixel-level correlations between paired query and support features implemented with the attention mechanism of Transformer. Nevertheless, these approaches suffer from heavy computation due to dot-product attention between all pixels of support and query features. In this paper, we propose a novel framework, termed ProtoFormer, built upon the Transformer architecture, to fully capture spatial details in query features. ProtoFormer treats the abstracted prototype of the target class in support features as the Query and the query features as Key and Value embeddings, which are input to the Transformer decoder. This approach enables better capture of spatial details and focuses on the semantic features of the target class in the query image. The output of the Transformer-based module can be interpreted as semantic-aware dynamic kernels that filter the segmentation mask from the enriched query features. Extensive experiments conducted on PASCAL-5i\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$5^{i}$$\\end{document} and COCO-20i\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$20^{i}$$\\end{document} datasets demonstrate that ProtoFormer significantly outperforms the state-of-the-art methods in FSS.

TFAN: A Task-adaptive Feature Alignment Network for few-shot website fingerprinting attacks on Tor

Few-shot website fingerprinting (WF) attacks aim to infer which website a user browsed through anonymity networks, such as Tor, using limited labeled traces. Recent methods either adopt complex metric strategies or perform time-consuming transfer learning, neither of which yields the most efficient performance in dynamic network environments. In this paper, we introduce a novel Task-adaptive Feature Alignment Network (TFAN) following the meta-learning paradigm. TFAN regards the few-shot WF attack as a feature alignment problem in class latent space, aiming to depict each location in the query feature map as a weighted sum of support features of a given class. Ridge regression provides a closed-form solution without extra parameters or techniques, ensuring high computational efficiency. Moreover, we also propose a Task-adaptive Modulation Unit (TMU), which activates the differences between support prototypes to generate task-level channel weights, making channels with significant discriminative details for each task contribute more to alignment. Extensive experiments on public Tor datasets demonstrate the superiority of TFAN in different scenarios. Notably, it is the only method that maintains over 90% accuracy in the 1-shot setting even 42 days later. Our code is available at https://github.com/Crybaby98/TFAN.

Few-shot fine-grained recognition in remote sensing ship images with global and local feature aggregation

Remote sensing ship image detection methods have broad application prospects in areas such as maritime traffic and fisheries management. However, previous detection methods relied heavily on a large amount of accurately annotated training data. When the number of remote sensing ship targets is scarce, the detection performance of previous methods is unsatisfactory. To address this issue, this paper proposes a few-shot detection method based on global and local feature aggregation. Specifically, we introduce global and local feature aggregation. We aggregate query-image global and local features with support features. This encourages the model to learn invariant features under varying global feature conditions which enhances the model’s performance in training and inference. Building upon this, we propose combined feature aggregation, where query features are aggregated with all support features in the same batch, further reducing the confusion of target features caused by the imbalance between base-class samples and novel-class samples, improving the model’s learning effectiveness for novel classes. Additionally, we employ an adversarial autoencoder to reconstruct support features, enhancing the model’s generalization performance. Finally, the model underwent extensive experiments on the publicly available remote sensing ship dataset HRSC-2016. The results indicate that compared to the baseline model, our model achieved new state-of-the-art performance under various dataset settings. This model presented in this paper will provide new insights for few-shot detection work based on meta-learning.

Class-agnostic counting and localization with feature augmentation and scale-adaptive aggregation

Compared to traditional class-specific counting, class-agnostic counting (CAC) enables the counting of arbitrary novel categories with only a few exemplars. The core issue of CAC lies in feature matching between the query image and exemplars. However, the scarce user-given exemplars make it difficult to adequately cover the diverse appearance of target objects, which poses challenges for feature matching. In this paper, we propose a Class-Agnostic Counting and Localization model (CACL) to enrich the exemplars by mining more inherent instances from the query feature. Specifically, CACL consists mainly of Self-Augment Feature Matching (SAFM) and Scale-Adaptive Spatial Aggregation (SASA). SAFM collects features of target instances frequently occurring in the query feature to augment the limited user-given exemplar, thus producing more comprehensive semantic features of the target category. To improve the robustness against scale variation, we design the SASA to enable the model to aggregate multiscale features adaptively. Compared to previous density map-based approaches, our method can locate the target objects which is more practical in real-world applications. Extensive experiments on FSC-147 suggest that our method achieves a competitive counting performance compared to state-of-the-art methods. Moreover, cross-dataset experiments demonstrate superior generalization in class-specific counting tasks.

Authenticated Range Querying of Historical Blockchain Healthcare Data Using Authenticated Multi-Version Index

With growing adoption of blockchain in established and emerging applications, there is an increasing need to support efficient ad hoc querying of authenticated historical data. This is especially true in fields such as healthcare to meet the rigorous security and regulatory requirements of ever-expanding digital health platforms. Existing blockchain systems, however, offer little or no support for querying capabilities over historical data. Although a full blockchain archive node can be used to maintain historical records of all executed transactions on the chain, it is not scalable when dealing with large volumes of data. Moreover, such ‘offline’ historical data lack tamper evidence support. To address these issues, we introduce an authenticated index structure called Authenticated Multi-Version Skip List (AMVSL), designed to support a rich set of query features over historical blockchain data. We further present three range queries: SVRK, MVRK, and MVAK, which offer querying over a range of keys and a range of versions. Our experimental evaluation of two healthcare-inspired examples demonstrates that AMVSL efficiently supports these queries and can achieve performance that is several orders of magnitude faster than existing authenticated data structures.

FerroLigandDB: A Ferroptosis Ligand Database of Structure-Activity Relations.

Ferroptosis is an iron-dependent programmed cell death characterized by lipid peroxidation that is linked to the pathophysiological processes in many diseases, such as neurodegenerative diseases, cancers, ischemia-reperfusion injuries, and organ damages. Many proteins are associated with ferroptosis signal transduction pathways. Novel chemical compounds are demanded to explore and regulate these pathways. Therefore, a ferroptosis ligand database, which holds relations among chemical structures, targets, bioactivities, and diseases, is needed for discovering and designing new ferroptosis regulators. This work reports FerroLigandDB, a manually curated database for small-molecular ferroptosis regulators. The database comprises 466 ferroptosis inducer entries (with 380 unique molecular structures) and 539 ferroptosis inhibitor entries (with 468 unique molecular structures) (note: one compound can be recorded as multiple entries due to the different assays). Each ferroptosis ligand entry is detailed with compound IDs, structure attributes, bioactivity values, test objects, target information, associated diseases, and references. The fields in the FerroLigandDB database implicitly contain relationships among chemical structures, bioactivities, targets, and diseases. Thus, FerroLigandDB is a comprehensive resource for scientists to design and discover novel ferroptosis regulators. The user interface of FerroLigandDB is implemented with query features and data visualization facilities. With compound identifiers, the compounds are linked to the records of other chemoinformatics databases (such as PubChem and SciFinder). The FerroLigandDB database is freely accessible at http://ferr.gulab.org.cn/.

Query-Oriented Micro-Video Summarization.

Query-oriented micro-video summarization task aims to generate a concise sentence with two properties: (a) summarizing the main semantic of the micro-video and (b) being expressed in the form of search queries to facilitate retrieval. Despite its enormous application value in the retrieval area, this direction has barely been explored. Previous studies of summarization mostly focus on the content summarization for traditional long videos. Directly applying these studies is prone to gain unsatisfactory results because of the unique features of micro-videos and queries: diverse entities and complex scenes within a short time, semantic gaps between modalities, and various queries in distinct expressions. To specifically adapt to these characteristics, we propose a query-oriented micro-video summarization model, dubbed QMS. It employs an encoder-decoder-based transformer architecture as the skeleton. The multi-modal (visual and textual) signals are passed through two modal-specific encoders to obtain their representations, followed by an entity-aware representation learning module to identify and highlight critical entity information. As to the optimization, regarding the large semantic gaps between modalities, we assign different confidence scores according to their semantic relevance in the optimization process. Additionally, we develop a novel strategy to sample the effective target query among the diverse query set with various expressions. Extensive experiments demonstrate the superiority of the QMS scheme, on both the summarization and retrieval tasks, over several state-of-the-art methods.

Self-support matching networks with multiscale attention for few-shot semantic segmentation

Recent advancements in Few-shot segmentation (FSS) have displayed remarkable capabilities in predicting segmentation masks for unseen class images, using only a limited number of annotated images. However, existing methods have overlooked the influence of contextual information on segmentation and primarily rely on supporting prototypes, with limited research on query prototyping. Effectively utilizing multi-scale features and query information poses a challenging problem in this domain. To address these challenges, this paper proposes a novel approach called the multi-scale and attention-based self-support prototype few-shot semantic segmentation network (MASNet). First, a multi-scale feature enhancement module is designed to obtain features at different scales to enrich global context information. Then, simple and efficient channel attention is utilized to guide the query features related to the target class. Finally, the query prototype is matched with the query features using a self-supporting matching module. This strategy efficiently captures class-based features and addresses the issue of intra-class variance in few-shot segmentation. The experimental results on Pascal-5i, COCO-20i and Abdominal MRI datasets demonstrate that the proposed method achieves remarkable robustness and improved accuracy performance.

CompleteDT: Point cloud completion with information-perception transformers

In this work, we propose a novel point cloud completion network, called CompleteDT. To fully capture the 3D geometric structure of point clouds, we introduce an Information-Perception Transformer (IPT) that can simultaneously capture local features and global geometric relations. CompleteDT comprises a Feature Encoder, Query Generator, and Query Decoder. Feature Encoder extracts local features from multi-resolution point clouds to capture intricate geometrical structures. Query Generator uses the proposed IPT, utilizing the Point Local Attention (PLA) and Point Global Attention (PGA) modules, to learn local features and global correlations, and generate query features that represent predicted point clouds. The PLA captures local information within local points by adaptively measuring weights of neighboring points, while PGA adapts multi-head self-attention by transforming it into a layer-by-layer form where each head learns global features in a high-dimensional space of different dimensions. By dense connections, the module allows for direct information exchange between each head and facilitates the capture of long global correlations. By combining the strengths of both PLA and PGA, the IPT can fully leverage local and global features to facilitate CompleteDT to complete point clouds. Lastly, the query features undergo refining to generate a complete point cloud through the Query Decoder. Our experimental results demonstrate that CompleteDT outperforms current state-of-the-art methods, effectively learning from incomplete inputs and predicting complete outputs.

CSTrans: Correlation-guided Self-Activation Transformer for Counting Everything

Counting everything, also named few-shot counting, requires a model to be able to count objects with any novel (unseen) category giving few exemplar boxes. However, the existing few-shot counting methods are sub-optimal due to weak feature representation, such as the correlation between the exemplar patch and query feature, and contextual dependencies in density map prediction. In this paper, we propose a very simple but effective method, CSTrans, consisting of a Correlation-guided Self-Activation (CSA) module and a Local Dependency Transformer (LDT) module, to mitigate the above two issues, respectively. The CSA utilizes the correlation map to activate the semantic features and suppress the noisy influence of the query features, aiming at mining the potential relation while enriching correlation representation. Furthermore, the LDT incorporates a Transformer to explore local contextual dependencies and predict the density map. Our method achieves competitive performance on FSC-147 and CARPK datasets. We hope its simple implementation and superior performance can serve as a new and strong baseline for few-shot counting tasks and attract more interest in designing simple but effective models in future studies. Our code for CSTrans is available at https://github.com/gaobb/CSTrans.

A hybrid network of NARX and DS-attention applied for the state estimation of lithium-ion batteries

Monitoring the state of the battery, including the state of charge (SOC) and state of health (SOH), is crucial for ensuring the safety and reliability of electrical equipment. The paper presents a novel hybrid network that combines nonlinear autoregressive model with exogenous inputs (NARX) and DS-attention. The proposed DS-attention method establishes a robust mapping relationship between inputs and outputs, it is a specialized method of the recurrent neural network that enhances the estimation performance by incorporating division function and self-adaptive function into the attention mechanism. The division function is designed to efficiently differentiate between exogenous inputs and state outputs, thereby minimizing the potential for cross-interference between them, and the self-adaptive function optimizes the query features within the attention mechanism. The results demonstrate that the hybrid method of NARX and DS-attention achieves higher prediction accuracy for SOH and SOC estimation of lithium-ion batteries. Specifically, compared with the best-performing similar methods, the proposed hybrid method enhanced the prediction accuracy by 22.9%, 60.7%, and 51.2% across three different SOH datasets, respectively. In terms of long sequence SOC forecasting, the improvements in prediction accuracy under two different working conditions are 82.5% and 60.5%, respectively. The proposed algorithm optimizes the attention mechanism based on the characteristics of the NARX, resulting in higher estimation accuracy for predicting the state of lithium batteries.