{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,4]],"date-time":"2026-09-04T14:15:16Z","timestamp":1788531316637,"version":"build-2803163510"},"reference-count":60,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,9,4]],"date-time":"2026-09-04T00:00:00Z","timestamp":1788480000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100007620","name":"Department of Education of Liaoning Province","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100007620","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Psychiatry"],"abstract":"<jats:p>Facial Expression Recognition (FER) has emerged as an important affective phenotyping tool in computational psychiatry, providing objective measurements of facial expressivity associated with psychomotor abnormalities in psychiatric disorders. From the perspective of smart product design, deploying deep learning frameworks for longitudinal psychiatric monitoring requires balancing recognition accuracy with hardware constraints. Since existing dual-stream networks often rely on computationally intensive global cross-attention, their integration into resourceconstrained products or interactive terminals remains challenging. To address these limitations, we present KGLA-Net, a dual-stream network. Our framework extracts parallel representations through a landmark backbone and an image backbone across three hierarchical stages. In the structural stream, a feature refinement module processes emotion-relevant cues, which are then aggregated through max-pooling to form multi-scale refined keypoint features. This structural prior drives a Keypoint-Guided Local Aggregation (KGLA) module, employing multi-head cross attention across localized multi-scale windows (8 \u00d7 8, 4 \u00d7 4, and 2 \u00d7 2) to guide the aggregation of image features, thereby reducing the computational complexity to linear complexity. Finally, the aggregated features are processed via keypoint feature self-attention before interacting with a zero-initialized query in stacked attention blocks. Experimental evaluations demonstrate that KGLA-Net achieves competitive FER performance while maintaining high computational efficiency. The proposed framework provides an efficient solution for facial affective computing and offers a lightweight foundation for future research on computational psychiatry.<\/jats:p>","DOI":"10.3389\/fpsyt.2026.1903055","type":"journal-article","created":{"date-parts":[[2026,9,4]],"date-time":"2026-09-04T13:23:22Z","timestamp":1788528202000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["KGLA-Net: an efficient facial affective computing framework for psychomotor phenotyping in psychiatry"],"prefix":"10.3389","volume":"17","author":[{"given":"Fengze","family":"Li","sequence":"first","affiliation":[{"name":"School of Industrial Design, Lu Xun Academy of Fine Arts","place":["Liaoning, Shenyang, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jiaying","family":"Yu","sequence":"additional","affiliation":[{"name":"School of Industrial Design, Lu Xun Academy of Fine Arts","place":["Liaoning, Shenyang, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Minglei","family":"Sun","sequence":"additional","affiliation":[{"name":"School of Industrial Design, Lu Xun Academy of Fine Arts","place":["Liaoning, Shenyang, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xiang","family":"Li","sequence":"additional","affiliation":[{"name":"School of Industrial Design, Lu Xun Academy of Fine Arts","place":["Liaoning, Shenyang, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,9,4]]},"reference":[{"key":"B1","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR.2017.277","article-title":"Reliable crowdsourcing and deep locality-preserving learning for expression recognition in the wild","author":"Li","year":"2017"},{"key":"B2","doi-asserted-by":"publisher","first-page":"18","DOI":"10.1109\/taffc.2017.2740923","article-title":"Affectnet: A database for facial expression, valence, and arousal computing in the wild","volume":"10","author":"Mollahosseini","year":"2017","journal-title":"IEEE Trans Affect Comput"},{"key":"B3","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR46437.2021.00618","article-title":"Dive into ambiguity: Latent distribution mining and pairwise uncertainty estimation for facial expression recognition","author":"She","year":"2021"},{"key":"B4","doi-asserted-by":"publisher","DOI":"10.1016\/j.inffus.2025.102971","article-title":"Explainable multi-frequency and multi-region fusion model for affective brain-computer interfaces","volume":"118","author":"Wang","year":"2025","journal-title":"Inf Fusion"},{"key":"B5","doi-asserted-by":"publisher","DOI":"10.1109\/taffc.2026.3675425","article-title":"Explainable affective body expression recognition with multi-scale spatiotemporal encoding and llm-based reasoning","author":"Wang","year":"2026","journal-title":"IEEE Trans Affect Comput"},{"key":"B6","doi-asserted-by":"crossref","DOI":"10.1609\/aaai.v37i2.25272","article-title":"See your emotion from gait using unlabeled skeleton data","author":"Lu","year":"2023"},{"key":"B7","doi-asserted-by":"crossref","DOI":"10.1109\/ICCV48922.2021.00358","article-title":"Transfer learning relation-aware facial expression representations with transformers","author":"Xue","year":"2021"},{"key":"B8","article-title":"Distract your attention: Multi-head cross attention network for facial expression recognition","author":"Wen","year":"2021","journal-title":"Arxiv"},{"key":"B9","doi-asserted-by":"crossref","DOI":"10.1109\/ICCVW60793.2023.00339","article-title":"Poster: A pyramid cross-fusion transformer network for facial expression recognition","author":"Zheng","year":"2023"},{"key":"B10","doi-asserted-by":"publisher","DOI":"10.1016\/j.patcog.2024.110951","article-title":"Poster++: A simpler and stronger facial expression recognition network","volume":"157","author":"Mao","year":"2025","journal-title":"Pattern Recognit"},{"key":"B11","article-title":"An image is worth 16x16 words: Transformers for image recognition at scale","author":"Dosovitskiy","year":"2020","journal-title":"Arxiv"},{"key":"B12","doi-asserted-by":"publisher","DOI":"10.1109\/tpami.2007.1110","article-title":"Dynamic texture recognition using local binary patterns with an application to facial expressions","volume":"29","author":"Zhao","year":"2007","journal-title":"IEEE Trans Pattern Anal Mach Intell"},{"key":"B13","first-page":"1","article-title":"Discriminative deep feature learning for facial emotion recognition","volume-title":"1st International Conference on Multimedia Analysis and Pattern Recognition (MAPR)","author":"Sang","year":"2018"},{"key":"B14","doi-asserted-by":"publisher","DOI":"10.1007\/s00521-020-05676-y","article-title":"Fer-net: facial expression recognition using deep neural net","volume":"33","author":"Mohan","year":"2021","journal-title":"Neural Comput Appl"},{"key":"B15","doi-asserted-by":"publisher","DOI":"10.3390\/s22103729","article-title":"Facial expression recognition based on squeeze vision transformer","volume":"22","author":"Kim","year":"2022","journal-title":"Sensors"},{"key":"B16","doi-asserted-by":"publisher","DOI":"10.1109\/jbhi.2022.3233597","article-title":"Epic: Emotion perception by spatio-temporal interaction context of gait","volume":"28","author":"Lu","year":"2023","journal-title":"IEEE J BioMed Health Inf"},{"key":"B17","doi-asserted-by":"publisher","DOI":"10.1109\/taffc.2024.3392904","article-title":"Mast-gcn: Multi-scale adaptive spatial-temporal graph convolutional network for eeg-based depression recognition","volume":"15","author":"Lu","year":"2024","journal-title":"IEEE Trans Affect Comput"},{"key":"B18","doi-asserted-by":"publisher","DOI":"10.1016\/j.asoc.2024.111762","article-title":"A gradual self distillation network with adaptive channel attention for facial expression recognition","volume":"161","author":"Zhang","year":"2024","journal-title":"Appl Soft Comput"},{"key":"B19","doi-asserted-by":"publisher","first-page":"397","DOI":"10.1177\/18761364251348382","article-title":"Fs-dsn: A feature segmentation-based dual-stream network for robust facial expression recognition in uncontrolled environments","volume":"17","author":"Guo","year":"2025","journal-title":"J Ambient Intell Smart Environ"},{"key":"B20","doi-asserted-by":"publisher","DOI":"10.1007\/s11042-025-20813-2","article-title":"Facial micro-expression recognition based on dual-stream fusion network","volume":"84","author":"Sun","year":"2025","journal-title":"Multimedia Tools Appl"},{"key":"B21","doi-asserted-by":"publisher","first-page":"311","DOI":"10.1007\/s11227-024-06810-6","article-title":"Tscanet: a two-stream context aggregation network for weakly-supervised temporal action localization","volume":"81","author":"Zhang","year":"2025","journal-title":"J Supercomputing"},{"key":"B22","doi-asserted-by":"publisher","first-page":"201","DOI":"10.1016\/j.patrec.2026.01.007","article-title":"Mm-net: Facial expression recognition based on multi-level and multi-scale attention mechanisms","author":"Wang","year":"2026","journal-title":"Pattern Recognit Lett"},{"key":"B23","doi-asserted-by":"publisher","DOI":"10.1109\/taffc.2022.3208309","article-title":"Flepnet: feature level ensemble parallel network for facial expression recognition","volume":"13","author":"Karnati","year":"2022","journal-title":"IEEE Trans Affect Comput"},{"key":"B24","doi-asserted-by":"crossref","DOI":"10.1609\/aaai.v39i2.32135","article-title":"Understanding emotional body expressions via large language models","author":"Lu","year":"2025"},{"key":"B25","doi-asserted-by":"crossref","DOI":"10.1109\/ICCV48922.2021.00060","article-title":"Tokens-to-token vit: Training vision transformers from scratch on imagenet","author":"Yuan","year":"2021"},{"key":"B26","doi-asserted-by":"crossref","DOI":"10.1109\/ICCV48922.2021.00041","article-title":"Crossvit: Cross-attention multi-scale vision transformer for image classification","author":"Chen","year":"2021"},{"key":"B27","doi-asserted-by":"crossref","DOI":"10.1109\/ICCV48922.2021.00061","article-title":"Pyramid vision transformer: A versatile backbone for dense prediction without convolutions","author":"Wang","year":"2021"},{"key":"B28","doi-asserted-by":"crossref","DOI":"10.1109\/ICCV48922.2021.00986","article-title":"Swin transformer: Hierarchical vision transformer using shifted windows","author":"Liu","year":"2021"},{"key":"B29","article-title":"Vision mamba: Efficient visual representation learning with bidirectional state space model","author":"Zhu","year":"2024"},{"key":"B30","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR52734.2025.01390","article-title":"Efficientvim: Efficient vision mamba with hidden state mixer based state space duality","author":"Lee","year":"2025"},{"key":"B31","doi-asserted-by":"publisher","first-page":"13826","DOI":"10.1038\/s41598-026-44396-6","article-title":"Fermam: a lightweight dual-source and multi-scale fusion framework for facial expression recognition","volume":"16","author":"Gao","year":"2026","journal-title":"Sci Rep"},{"key":"B32","doi-asserted-by":"publisher","first-page":"331","DOI":"10.1007\/s00530-026-02388-6","article-title":"Mambaer: Mamba with knowledge-learning hierarchical attention for facial expression recognition","volume":"32","author":"Pei","year":"2026","journal-title":"Multimedia Syst"},{"key":"B33","doi-asserted-by":"crossref","DOI":"10.1109\/ICCV.2019.01024","article-title":"Context-aware emotion recognition networks","author":"Lee","year":"2019"},{"key":"B34","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR.2019.00482","article-title":"Arcface: Additive angular margin loss for deep face recognition","author":"Deng","year":"2019"},{"key":"B35","unstructured":"Chen\n              C\n            \n          \n          Pytorch Face Landmark: A Fast and Accurate Facial Landmark Detector\n          \n          2021"},{"key":"B36","doi-asserted-by":"publisher","DOI":"10.1109\/access.2020.3010018","article-title":"Pyramid with super resolution for in-the-wild facial expression recognition","volume":"8","author":"Vo","year":"2020","journal-title":"IEEE Access"},{"key":"B37","doi-asserted-by":"crossref","DOI":"10.1109\/WACV48630.2021.00245","article-title":"Facial expression recognition in the wild via deep attentive center loss","author":"Farzaneh","year":"2021"},{"key":"B38","doi-asserted-by":"publisher","DOI":"10.1109\/taffc.2021.3122146","article-title":"Facial expression recognition with visual transformers and attentional selective fusion","volume":"14","author":"Ma","year":"2021","journal-title":"IEEE Trans Affect Comput"},{"key":"B39","article-title":"Learning to amend facial expression representation via de-albino and affinity","author":"Shi","year":"2021","journal-title":"Arxiv"},{"key":"B40","doi-asserted-by":"crossref","DOI":"10.1609\/aaai.v35i4.16465","article-title":"Robust lightweight facial expression recognition network with label distribution training","author":"Zhao","year":"2021"},{"key":"B41","doi-asserted-by":"publisher","DOI":"10.1109\/tip.2021.3093397","article-title":"Learning deep global multi-scale and local attention features for facial expression recognition in the wild","volume":"30","author":"Zhao","year":"2021","journal-title":"IEEE Trans Image Process"},{"key":"B42","doi-asserted-by":"publisher","DOI":"10.1016\/j.patcog.2022.108737","article-title":"Improving the facial expression recognition and its interpretability via generating expression pattern-map","volume":"129","author":"Zhang","year":"2022","journal-title":"Pattern Recognit"},{"key":"B43","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR52688.2022.01965","article-title":"Face2exp: Combating data biases for facial expression recognition","author":"Zeng","year":"2022"},{"key":"B44","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-031-19809-0_24","article-title":"Learn from all: Erasing attention consistency for noisy label facial expression recognition","volume-title":"European Conference on Computer Vision","author":"Zhang","year":"2022"},{"key":"B45","doi-asserted-by":"publisher","DOI":"10.1109\/tcsvt.2022.3165321","article-title":"Adaptive multilayer perceptual attention network for facial expression recognition","volume":"32","author":"Liu","year":"2022","journal-title":"IEEE Trans Circuits Syst Video Technol"},{"key":"B46","doi-asserted-by":"publisher","first-page":"110530","DOI":"10.2139\/ssrn.4267797","article-title":"Fer-chc: Facial expression recognition with cross-hierarchy contrast","volume":"145","author":"Wu","year":"2023","journal-title":"Appl Soft Comput"},{"key":"B47","doi-asserted-by":"publisher","DOI":"10.1109\/access.2023.3237817","article-title":"Fine-tuning swin transformer and multiple weights optimality seeking for facial expression recognition","volume":"11","author":"Feng","year":"2023","journal-title":"IEEE Access"},{"key":"B48","doi-asserted-by":"publisher","DOI":"10.1109\/tcsvt.2023.3304724","article-title":"Swinface: A multi-task transformer for face recognition, expression recognition, age estimation and attribute estimation","volume":"34","author":"Qin","year":"2023","journal-title":"IEEE Trans Circuits Syst Video Technol"},{"key":"B49","first-page":"1","article-title":"Multi-branch attention consistency network for facial expression recognition","volume-title":"IECON 2023-49th Annual Conference of the IEEE Industrial Electronics Society","author":"Ji","year":"2023"},{"key":"B50","doi-asserted-by":"publisher","DOI":"10.1016\/j.ins.2023.119640","article-title":"Face2nodes: Learning facial expression representations with relation-aware dynamic graph convolution networks","volume":"649","author":"Jiang","year":"2023","journal-title":"Inf Sci"},{"key":"B51","doi-asserted-by":"publisher","DOI":"10.1109\/tcsvt.2023.3234312","article-title":"Multi-relations aware network for in-the-wild facial expression recognition","volume":"33","author":"Chen","year":"2023","journal-title":"IEEE Trans Circuits Syst Video Technol"},{"key":"B52","doi-asserted-by":"publisher","DOI":"10.1016\/j.knosys.2023.110451","article-title":"Squeezexpnet: dual-stage convolutional neural network for accurate facial expression recognition with attention mechanism","volume":"269","author":"Shahid","year":"2023","journal-title":"Knowledge-Based Syst"},{"key":"B53","doi-asserted-by":"publisher","DOI":"10.1016\/j.neunet.2022.11.025","article-title":"Ssa-icl: multi-domain adaptive attention with intra-dataset continual learning for facial expression recognition","volume":"158","author":"Gao","year":"2023","journal-title":"Neural Networks"},{"key":"B54","doi-asserted-by":"publisher","DOI":"10.1016\/j.neunet.2023.11.033","article-title":"Hierarchical attention network with progressive feature fusion for facial expression recognition","volume":"170","author":"Tao","year":"2024","journal-title":"Neural Networks"},{"key":"B55","doi-asserted-by":"publisher","first-page":"110311","DOI":"10.1016\/j.patcog.2024.110311","article-title":"Pose-robust personalized facial expression recognition through unsupervised multi-source domain adaptation","volume":"150","author":"Wang","year":"2024","journal-title":"Pattern Recognit"},{"key":"B56","doi-asserted-by":"crossref","DOI":"10.1145\/3242969.3264978","article-title":"Video-based emotion recognition using deeply supervised neural networks","author":"Fan","year":"2018"},{"key":"B57","doi-asserted-by":"publisher","DOI":"10.1109\/access.2020.3048693","article-title":"Graph reasoning-based emotion recognition network","volume":"9","author":"Gao","year":"2021","journal-title":"IEEE Access"},{"key":"B58","doi-asserted-by":"publisher","DOI":"10.1007\/s00521-021-06778-x","article-title":"Global-local attention for emotion recognition","volume":"34","author":"Le","year":"2022","journal-title":"Neural Comput Appl"},{"key":"B59","doi-asserted-by":"publisher","DOI":"10.3390\/sym15061228","article-title":"Symmetric multi-scale residual network ensemble with weighted evidence fusion strategy for facial expression recognition","volume":"15","author":"Liu","year":"2023","journal-title":"Symmetry"},{"key":"B60","doi-asserted-by":"publisher","DOI":"10.1016\/j.ins.2022.11.068","article-title":"Patch attention convolutional vision transformer for facial expression recognition with occlusion","volume":"619","author":"Liu","year":"2023","journal-title":"Inf Sci"}],"container-title":["Frontiers in Psychiatry"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fpsyt.2026.1903055\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,4]],"date-time":"2026-09-04T13:23:43Z","timestamp":1788528223000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fpsyt.2026.1903055\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,4]]},"references-count":60,"alternative-id":["10.3389\/fpsyt.2026.1903055"],"URL":"https:\/\/doi.org\/10.3389\/fpsyt.2026.1903055","relation":{},"ISSN":["1664-0640"],"issn-type":[{"value":"1664-0640","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,4]]},"article-number":"1903055"}}