IEEE Transactions on Cognitive and Developmental Systems最新文献

{"title":"GTR: Global Token Retrieval for Window-Based Vision Transformer","authors":"Xin Huang;Long Chen;Qian Yu;Lei Qi","doi":"10.1109/TCDS.2026.3668720","DOIUrl":"https://doi.org/10.1109/TCDS.2026.3668720","url":null,"abstract":"Vision transformers (ViTs) have achieved remarkable success in visual recognition tasks due to their strong capability to model long-range dependencies through global self-attention. However, the inherent quadratic complexity of global attention severely limits their scalability when processing high-resolution inputs. To address this issue, window-based attention mechanisms have been proposed to reduce computational overhead, but they often struggle with limited receptive fields and insufficient global context modeling. In this work, we propose global token retrieval vision transformer (GTR-ViT), a novel and efficient ViT framework that integrates a global token retrieval (GTR) mechanism with an enhanced feed-forward network (FFN) in a synergistic manner. The GTR mechanism dynamically retrieves globally relevant tokens for each token, enriching the local features with comprehensive global context. These globally enhanced features are then fed into a gated feed-forward network equipped with depthwise convolution, which further refines local representations through adaptive feature modulation. In turn, the improved local features serve as more representative queries for the GTR module, enabling more accurate GTR in subsequent layers. This mutual reinforcement forms a virtuous cycle of local-global interaction, significantly boosting the model’s capability in both context understanding and feature discrimination. Extensive experiments on image classification and downstream tasks across multiple datasets demonstrate that GTR-ViT consistently surpasses prior state-of-the-art methods, highlighting the effectiveness of this collaborative design.","PeriodicalId":54300,"journal":{"name":"IEEE Transactions on Cognitive and Developmental Systems","volume":"18 5","pages":"1187-1199"},"PeriodicalIF":4.7,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149176334","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"RBA-FE: A Robust Brain-Inspired Audio Feature Extractor for Depression Diagnosis","authors":"Yu-Xuan Wu;Ziyan Huang;Bin Hu;Zhi-Hong Guan","doi":"10.1109/TCDS.2026.3684337","DOIUrl":"https://doi.org/10.1109/TCDS.2026.3684337","url":null,"abstract":"This article proposes a robust brain-inspired audio feature extractor (RBA-FE) model for depression diagnosis, using an improved hierarchical network architecture. Most deep learning models achieve state-of-the-art performance for image-based diagnostic tasks, ignoring the counterpart audio features. To tailor the noise challenge, RBA-FE leverages six acoustic features extracted from the raw audio, capturing both spatial characteristics and temporal dependencies. This hybrid attribute helps alleviate the precision limitation in audio feature extraction within other learning models like deep residual shrinkage networks. To deal with the noise issues, our model incorporates an improved spiking neuron model, called adaptive rate smooth leaky integrate-and-fire (ARSLIF). The ARSLIF model emulates the mechanism of “retuning of cellular signal selectivity” in the brain attention systems, which enhances the model robustness against environmental noises in audio data. Experimental results demonstrate that RBA-FE achieves state-of-the-art accuracy on the MODMA dataset, respectively with 0.8750, 0.8974, 0.8750, and 0.8750 in precision, accuracy, recall, and F1 score. Extensive experiments on the AVEC2014 and DAIC-WOZ datasets both show enhancements in noise robustness. It is further indicated by comparison that the ARSLIF neuron model suggests the abnormal firing pattern within the feature extraction on depressive audio data, offering brain-inspired interpretabilit.","PeriodicalId":54300,"journal":{"name":"IEEE Transactions on Cognitive and Developmental Systems","volume":"18 5","pages":"1369-1383"},"PeriodicalIF":4.7,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149176100","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multimodal EEG-EMG Deep Learning Framework for Motor Imagery Classification While Using a Robotic Hand Exoskeleton","authors":"Sushant Kumar;Sushant Maurya;Ashish Dutta","doi":"10.1109/TCDS.2026.3682048","DOIUrl":"https://doi.org/10.1109/TCDS.2026.3682048","url":null,"abstract":"A brain-computer interface (BCI) is a revolutionary technology with the potential to transform assistive applications for people with disabilities and enhance human capabilities. However, the performance of electroencephalogram (EEG)-based motor imagery (MI) signals, commonly used in BCI systems, is often limited by low classification accuracy and poor adaptability. Similarly, electromyography (EMG) signals, widely used in motor rehabilitation, suffer from instability due to muscle fatigue. To improve the accuracy and stability of interactive training in a robotic hand exoskeleton system, we propose a new approach called multimodal vision transformer-based dilated temporal convolutional network (MViT-DTCNet) for classification of MI. The multimodal vision transformer captures detailed, fine-grained spatial features, while the dilated temporal convolutional network learns long-term temporal dependencies across both EEG and EMG signals using dilated convolutions at varying rates to preserve temporal resolution and multiscale patterns. By fusing EEG and EMG data at an intermediate stage, our approach significantly enhances the decoding of user intent. The experimental results with healthy participants demonstrate a classification accuracy of 96.40% ± 2.64 and a Cohen’s kappa of 0.93 ± 0.05 while using robotic hand exoskeletons, indicating stable and reliable performance of the proposed model for the advancement of assistive technologies and neurorehabilitation.","PeriodicalId":54300,"journal":{"name":"IEEE Transactions on Cognitive and Developmental Systems","volume":"18 5","pages":"1315-1324"},"PeriodicalIF":4.7,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149176238","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"CLAIM: Mitigating Catastrophic Forgetting in Continual Instruction Fine-Tuning Large Language Models","authors":"Xihe Qiu;Shaojie Shi;Teqi Hao;Xiaoyu Tan","doi":"10.1109/TCDS.2026.3674105","DOIUrl":"https://doi.org/10.1109/TCDS.2026.3674105","url":null,"abstract":"In recent times, large language models (LLMs) have garnered substantial interest from the research community, underpinned by their widespread applicability in numerous language-oriented tasks and their notable performance enhancement through instruction fine-tuning techniques, grounded in human preference. However, enhancing the instruction-following proficiency of LLMs in specific tasks necessitates users undertaking supervised fine-tuning on already fine-tuned LLMs, leading to a substantial loss in generality and inducing catastrophic forgetting. This phenomenon significantly undermines the models’ generalization capacity across diverse tasks and intensifies hallucination in LLMs’ generation. Given the proprietary nature of high-quality instruction-following data, there’s a pressing need for methods that combat both catastrophic forgetting and hallucination without relying on prior fine-tuning datasets. We introduce <sc>CLAIM</small>: continual instruction fine-tuning large language models with augmented instruction self-training and Model averaging, a novel framework that leverages augmented instruction self-training and model averaging to counter these challenges. Empirical experiment results confirm <sc>CLAIM</small>’s ability to significantly curtail catastrophic forgetting and hallucinations during continual fine-tuning without accessing previously used instruction data.","PeriodicalId":54300,"journal":{"name":"IEEE Transactions on Cognitive and Developmental Systems","volume":"18 5","pages":"1256-1266"},"PeriodicalIF":4.7,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149175982","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Development and Preliminary Validation of the MODD System: A Sensorized $3times 3$ Platform for Cognitive-Motor Integration Assessment","authors":"Yuan Sun;Xuan Lin;Darkhan Baizhan;Gurmehar Lugani;Emerson Sebastião;Charles H. Hillman;Sean P. Mullen","doi":"10.1109/TCDS.2026.3690705","DOIUrl":"https://doi.org/10.1109/TCDS.2026.3690705","url":null,"abstract":"Cognitive-motor integration is a foundational feature of human behavior, enabling real-time coordination of perception, decision-making, and movement across diverse environments. This article introduces the motion detection device (MODD), a programmable <inline-formula> <tex-math>$mathbf{3times 3}$</tex-math> </inline-formula> grid platform designed to assess and train cognitive-motor performance through ecologically valid scenarios. Uniquely integrating pressure-sensitive input and embedded cognitive stimuli, the MODD enables dynamic measurement of integrated executive function, working memory, and agility—capabilities not replicated by existing commercial or standardized tools. In this preliminary validation study with 10 healthy young adults, MODD task performance showed consistent associations with well-established cognitive assessments (e.g., the Stroop Task, including metrics such as Stroop reaction time congruent, and the N-back task) and motor agility tests (e.g., the Illinois agility test and the spider agility drill). Bayesian estimation supported the <italic>convergent validity</i> of MODD measures relative to reference lab and field tests. Specifically, metrics of the MODD memory task (e.g., memory task accuracy) correlated with outcomes of the Sternberg task (e.g., Sternberg-5 accuracy match), while metrics of the MODD Go–NoGo Task (e.g., Go-NoGo Percent Interference) showed meaningful associations with Stroop Task results. Additionally, the MODD X-Task’s duration correlated with outcomes of standardized field agility tests. These results support the validation of our MODD battery as a highly precise assessment tool for real-time behavioral data and performance modeling, which also provides evidence for the construct validity of this tool. The versatility of the architecture of MODD allows for experimental control of stimulus response mapping and dual task demand control, thereby filling the void of valid and ecologically valid assessment tools in the area of cognitive-motor measurement. The development of an ecologically valid assessment tool for the study of human performance has broad applications in the advancement of human performance science and the area of adaptive rehabilitation.","PeriodicalId":54300,"journal":{"name":"IEEE Transactions on Cognitive and Developmental Systems","volume":"18 5","pages":"1200-1211"},"PeriodicalIF":4.7,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11509371","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149176085","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Brain–Machine Cross-Modal Enhancement Method With Deep Supervision for Visual Classification","authors":"Dongjun Liu;Weichen Dai;Honggang Liu;Hangjie Yi;Xuanyu Jin;Wanzeng Kong","doi":"10.1109/TCDS.2026.3667394","DOIUrl":"https://doi.org/10.1109/TCDS.2026.3667394","url":null,"abstract":"Visual classification has achieved remarkable progress with the development of deep neural networks (DNNs). However, the reliance on large-scale labeled training sets limits the generalized classification ability of such models. To address this challenge, we propose a brain–machine cross-modal enhancement framework that leverages brain signals as auxiliary supervision during training. Unlike existing methods that impose brain signal constraints only at the final classification layer, our method introduces deep guidance through hidden-layer constraints, enabling better feature discrimination and improved knowledge transfer. Specifically, we adopt contrastive learning to align brain signal and image modalities in a shared embedding space, where brain signal-image pairs serve as cross-modal supervision. The image encoder is partitioned into multiple shallow components, each equipped with a dedicated classifier trained using soft targets distilled from the brain signal-image coupling space. This full-process guidance design facilitates the extraction of semantically rich features across layers and enhances model generalization. Experimental results demonstrate that our method outperforms previous brain–machine enhancement approaches. Visualization results further reveal that our method enables the visual model to focus on semantically relevant regions in a human-like manner. These findings demonstrate that brain signals can effectively enhance visual learning, offering a new perspective for brain-inspired artificial intelligence.","PeriodicalId":54300,"journal":{"name":"IEEE Transactions on Cognitive and Developmental Systems","volume":"18 5","pages":"1162-1173"},"PeriodicalIF":4.7,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149176012","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Surrogate-Assisted Neural Architecture Search With Performance Clustering","authors":"Yu Xue;Weilin Fang;Peng Chen;Mohamed Wahib;Yiyu Tan","doi":"10.1109/TCDS.2026.3672588","DOIUrl":"https://doi.org/10.1109/TCDS.2026.3672588","url":null,"abstract":"Evolutionary neural architecture search (ENAS) automates the design of neural network architectures but incurs significant computational costs during the evaluation phase of candidate neural networks in the evolutionary algorithm. To address this issue, surrogate models are widely adopted to estimate the performance of candidate architectures, thereby reducing resource consumption. However, existing surrogate models are primarily based on regression or classification. They are often hindered by interference from low-quality architectures, which limits search efficiency and final performance. To overcome these challenges, we propose surrogate-assisted neural architecture search with performance clustering (PCNAS). PCNAS introduces a contrastive learning-based performance clustering encoder (PCE). The PCE maps candidate architectures into an embedding space to form high-performance clusters while maximizing separation from low-performance architectures. This design mitigates the negative influence of poor-quality candidates. To enhance the discriminative power of the PCE, we incorporate a staged sample generation strategy to strengthen the hierarchical representation capacity of the PCE. Additionally, we introduce a training-free simple classifier to efficiently select high-performance architectures from the embedding space. PCNAS achieves high accuracy with low computational cost on NAS-Bench-101 and NAS-Bench-201, outperforming previous approaches. Furthermore, PCNAS exhibits strong efficiency in the DARTS search space, discovering a high-performing architecture with a 2.53% error rate using only 0.5 GPU days of search cost, and on MobileNetV3 with ImageNet, achieving 78.12% top-1 accuracy using only 0.3 GPU days.","PeriodicalId":54300,"journal":{"name":"IEEE Transactions on Cognitive and Developmental Systems","volume":"18 5","pages":"1241-1255"},"PeriodicalIF":4.7,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149176061","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"On Harnessing EEG Signals for the Comprehensive Assessment of Neurological Disorders: A Review","authors":"Zaib Unnisa;Asadullah Tariq;IrfanUd Din;Abdelkader Nasreddine Belkacem","doi":"10.1109/TCDS.2026.3673111","DOIUrl":"https://doi.org/10.1109/TCDS.2026.3673111","url":null,"abstract":"Electroencephalography (EEG) is a scientific technique used to analyze and decode brain activity to identify different neurodegenerative diseases and psychiatric disorders. The insufficiency of specialists and neurologists is attributed to the prevalence of intricate disorders. Prompt and accurate diagnosis is crucial in managing any healthcare issue. EEG signal has shown its value in the prompt identification and prediction of diseases, hence providing valuable assistance to specialists and neurologists. This systematic review examines and analyzes the use of EEG signals for the diagnosis of different disorders and the identification of their biomarkers, drawing on existing research. The research aims to examine the use of EEG signals in the diagnostic process and determine whether a disorder may be predicted using EEG signals in advance. This raises novel research inquiries, which are investigated via 100 academic publications. Findings: EEG signals have been shown to be effective not only for early-stage diagnosis but also for predicting diseases and disorders before they arise, such as Bipolar detection, cancer detection, and the prediction of harmful brain activities. The results indicate that EEG signals can provide novel prospects for more exploration, such as the prediction and prognosis of diseases.","PeriodicalId":54300,"journal":{"name":"IEEE Transactions on Cognitive and Developmental Systems","volume":"18 5","pages":"1128-1150"},"PeriodicalIF":4.7,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149176144","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Time Cell Inspired Temporal Codebook in Spiking Neural Networks for Enhanced Image Generation","authors":"Linghao Feng;Dongcheng Zhao;Sicheng Shen;Yiting Dong;Guobin Shen;Yi Zeng","doi":"10.1109/TCDS.2026.3666668","DOIUrl":"https://doi.org/10.1109/TCDS.2026.3666668","url":null,"abstract":"Spiking neural networks (SNNs) have garnered significant attention due to their biological plausibility and energy efficiency. However, their generative modeling capabilities remain relatively underexplored, particularly in the context of handling temporal information. In this work, we propose a novel SNN-based vector quantized-variational autoencoder (VQ-VAE) framework that incorporates a temporal codebook inspired by hippocampal time cells. The temporal codebook enables the model to effectively encode and utilize temporal dependencies, facilitating the generation of high-resolution, temporally consistent images and event-based data. Extensive experiments conducted on diverse datasets, including CelebA-HQ, LSUN Bedroom, LSUN Church, CIFAR-10, and N-MNIST, demonstrate the superior performance of the proposed framework. Our method achieves state-of-the-art results in image quality and temporal consistency, outperforming existing SNN-based generative models. Furthermore, through destructive experiments, we validate the critical role of the temporal codebook in enhancing temporal correlation and improving generative performance. In addition to its superior image generation capabilities, our model exhibits significantly lower energy consumption compared with ANN-based counterparts, leveraging the inherent sparsity of SNNs. These findings highlight the potential of integrating temporal dynamics into the latent space of generative models, paving the way for advancements in neuromorphic computing and biologically inspired machine learning.","PeriodicalId":54300,"journal":{"name":"IEEE Transactions on Cognitive and Developmental Systems","volume":"18 5","pages":"1151-1161"},"PeriodicalIF":4.7,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149176323","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"IEEE Transactions on Cognitive and Developmental Systems Information for Authors","authors":"","doi":"10.1109/TCDS.2026.3731641","DOIUrl":"https://doi.org/10.1109/TCDS.2026.3731641","url":null,"abstract":"","PeriodicalId":54300,"journal":{"name":"IEEE Transactions on Cognitive and Developmental Systems","volume":"18 5","pages":"C4-C4"},"PeriodicalIF":4.7,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11737405","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149176289","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}