{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,12]],"date-time":"2026-02-12T17:40:32Z","timestamp":1770918032112,"version":"3.50.1"},"reference-count":59,"publisher":"MDPI AG","issue":"10","license":[{"start":{"date-parts":[[2025,9,26]],"date-time":"2025-09-26T00:00:00Z","timestamp":1758844800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["61771347"],"award-info":[{"award-number":["61771347"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["www.mdpi.com"],"crossmark-restriction":true},"short-container-title":["Symmetry"],"abstract":"<jats:p>Facial beauty prediction (FBP) is a cutting-edge task in deep learning that aims to equip machines with the ability to assess facial attractiveness in a human-like manner. In human perception, facial beauty is strongly associated with facial symmetry, where balanced structures often reflect aesthetic appeal. Leveraging symmetry provides an interpretable prior for FBP and offers geometric constraints that enhance feature learning. However, existing multi-task FBP models still face challenges such as limited annotated data, insufficient frequency\u2013temporal modeling, and feature conflicts from task heterogeneity. The Mamba model excels in feature extraction and long-range dependency modeling but encounters difficulties in parameter sharing and computational efficiency in multi-task settings. In contrast, mixture-of-experts (MoE) enables adaptive expert selection, reducing redundancy while enhancing task specialization. This paper proposes MoMamba, a multi-task decoder combining Mamba\u2019s state-space modeling with MoE\u2019s dynamic routing to improve multi-scale feature fusion and adaptability. A detail enhancement module fuses high- and low-frequency components from discrete cosine transform with temporal features from Mamba, and a state-aware MoE module incorporates low-rank expert modeling and task-specific decoding. Experiments on SCUT-FBP and SCUT-FBP5500 demonstrate superior performance in both classification and regression, particularly in symmetry-related perception modeling.<\/jats:p>","DOI":"10.3390\/sym17101600","type":"journal-article","created":{"date-parts":[[2025,9,26]],"date-time":"2025-09-26T06:02:52Z","timestamp":1758866572000},"page":"1600","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":1,"title":["A Multi-Task Fusion Model Combining Mixture-of-Experts and Mamba for Facial Beauty Prediction"],"prefix":"10.3390","volume":"17","author":[{"given":"Junying","family":"Gan","sequence":"first","affiliation":[{"name":"School of Electronics and Information Engineering, Wuyi University, Jiangmen 529020, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-8058-2799","authenticated-orcid":false,"given":"Zhenxin","family":"Zhuang","sequence":"additional","affiliation":[{"name":"School of Electronics and Information Engineering, Wuyi University, Jiangmen 529020, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0007-0328-4511","authenticated-orcid":false,"given":"Hantian","family":"Chen","sequence":"additional","affiliation":[{"name":"School of Electronics and Information Engineering, Wuyi University, Jiangmen 529020, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Wenchao","family":"Xu","sequence":"additional","affiliation":[{"name":"School of Electronics and Information Engineering, Wuyi University, Jiangmen 529020, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Zhen","family":"Chen","sequence":"additional","affiliation":[{"name":"School of Electronics and Information Engineering, Wuyi University, Jiangmen 529020, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0000-3958-1576","authenticated-orcid":false,"given":"Huicong","family":"Li","sequence":"additional","affiliation":[{"name":"School of Electronic Information and Control Engineering, Guangzhou University of Software, Guangzhou 510990, China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2025,9,26]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"119","DOI":"10.1162\/089976606774841602","article-title":"Facial attractiveness: Beauty and the machine","volume":"18","author":"Eisenthal","year":"2006","journal-title":"Neural Comput."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"255","DOI":"10.1126\/science.aaa8415","article-title":"Machine learning: Trends, perspectives, and prospects","volume":"349","author":"Jordan","year":"2015","journal-title":"Science"},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"120286","DOI":"10.1016\/j.ins.2024.120286","article-title":"Outlier detection method based on high-density iteration","volume":"662","author":"Zhou","year":"2024","journal-title":"Inf. Sci."},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"103318","DOI":"10.1016\/j.ndteint.2024.103318","article-title":"A plane stress measurement method for CFRP material based on array LCR waves","volume":"151","author":"Sun","year":"2025","journal-title":"NDT E Int."},{"key":"ref_5","first-page":"1500","article-title":"A novel method to facial beauty prediction based on self-supervised learning","volume":"39","author":"Gan","year":"2023","journal-title":"Signal Process."},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"56892","DOI":"10.1109\/ACCESS.2020.2980248","article-title":"Asian female facial beauty prediction using deep neural networks via transfer learning and multi-channel feature fusion","volume":"8","author":"Zhai","year":"2020","journal-title":"IEEE Access"},{"key":"ref_7","doi-asserted-by":"crossref","unstructured":"Saeed, J.N., Abdulazeez, A.M., and Ibrahim, D.A. (2022, January 8). FIAC-Net: Facial image attractiveness classification based on light deep convolutional neural network. Proceedings of the 2022 Second International Conference on Computer Science, Engineering and Applications (ICCSEA), Gunupur, India.","DOI":"10.1109\/ICCSEA54677.2022.9936582"},{"key":"ref_8","first-page":"18100","article-title":"Card: Classification and regression diffusion models","volume":"35","author":"Han","year":"2022","journal-title":"Adv. Neural. Inf. Process Syst."},{"key":"ref_9","doi-asserted-by":"crossref","unstructured":"Boukhari, D.E., Chemsa, A., Taleb-Ahmed, A., Ajgou, R., and Bouzaher, M.T. (2023). Facial beauty prediction using an ensemble of deep convolutional neural networks. Eng. Proc., 56.","DOI":"10.3390\/ASEC2023-15400"},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"7269","DOI":"10.33022\/ijcs.v13i1.3743","article-title":"Facial Beauty Prediction Based on Deep Learning: A Review","volume":"13","author":"Arabo","year":"2024","journal-title":"Indones. J. Comput. Sci."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"123644","DOI":"10.1016\/j.eswa.2024.123644","article-title":"Learning feature alignment across attribute domains for improving facial beauty prediction","volume":"249","author":"Sun","year":"2024","journal-title":"Expert Syst. Appl."},{"key":"ref_12","doi-asserted-by":"crossref","unstructured":"Lin, L., Liang, L., Jin, L., and Chen, W. (2019, January 10\u201316). Attribute-Aware Convolutional Neural Networks for Facial Beauty Prediction. Proceedings of the 28th International Joint Conference on Artificial Intelligence (IJCAI), Macao, China.","DOI":"10.24963\/ijcai.2019\/119"},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"1228","DOI":"10.1109\/TCSVT.2023.3292995","article-title":"Multi-task learning with multi-query transformer for dense prediction","volume":"34","author":"Xu","year":"2023","journal-title":"IEEE Trans. Circuits Syst. Video Technol."},{"key":"ref_14","doi-asserted-by":"crossref","unstructured":"Shah, U., Tukur, M., Alzubaidi, M., Pintore, G., Gobbetti, E., Househ, M., Schneider, J., and Agus, M. (2024, January 17\u201318). MultiPanoWise: Holistic deep architecture for multi-task dense prediction from a single panoramic image. Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Vancouver, Canada.","DOI":"10.1109\/CVPRW63382.2024.00138"},{"key":"ref_15","unstructured":"Sirejiding, S., Bayramli, B., Lu, Y., Yang, Y., Alsarhan, T., Lu, H., and Ding, Y. (November, January 28). Task-Interaction-Free multi-task learning with efficient hierarchical feature representation. Proceedings of the 32nd ACM International Conference on Multimedia, Lisbon, Portugal."},{"key":"ref_16","unstructured":"Guo, P., Lee, C.Y., and Ulbricht, D. (2009, January 14\u201318). Learning to branch for multi-task learning. Proceedings of the International conference on machine learning (ICML), PMLR, Vienna, Austria."},{"key":"ref_17","first-page":"1","article-title":"A decoder-focused multitask network for semantic change detection","volume":"62","author":"Li","year":"2024","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_18","unstructured":"Lin, B., Jiang, W., Chen, P., Zhang, Y., Liu, S., and Chen, Y.C. (October, January 29). MTMamba: Enhancing multi-task dense scene understanding by mamba-based decoders. Proceedings of the European Conference on Computer Vision, Milan, Italy."},{"key":"ref_19","doi-asserted-by":"crossref","unstructured":"Chen, T., Chen, X., Du, X., Rashwan, A., Yang, F., Chen, H., Wang, Z., and Li, Y. (2023, January 1\u20136). AdaMV-MoE: Adaptive multi-task vision mixture-of-experts. Proceedings of the IEEE\/CVF International Conference on Computer Vision (ICCV), Paris, France.","DOI":"10.1109\/ICCV51070.2023.01591"},{"key":"ref_20","first-page":"28441","article-title":"M3vit: Mixture-of-experts vision transformer for efficient multi-task learning with model-accelerator co-design","volume":"35","author":"Fan","year":"2022","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_21","doi-asserted-by":"crossref","unstructured":"Ye, H.R., and Xu, D. (2023, January 1\u20136). Taskexpert: Dynamically assembling multi-task representations with memorial mixture-of-experts. Proceedings of the IEEE\/CVF International Conference on Computer Vision (ICCV), Paris, France.","DOI":"10.1109\/ICCV51070.2023.01995"},{"key":"ref_22","first-page":"35971","article-title":"On the parameterization and initialization of diagonal state space models","volume":"35","author":"Gu","year":"2022","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_23","first-page":"572","article-title":"Combining recurrent, convolutional, and continuous-time models with linear state space layers","volume":"34","author":"Gu","year":"2021","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_24","doi-asserted-by":"crossref","unstructured":"Peng, Z. (2024, January 21\u201325). Ptm-mamba: A ptm-aware protein language model with bidirectional gated mamba blocks. Proceedings of the 33rd ACM International Conference on Information and Knowledge Management (CIKM), Birmingham, UK.","DOI":"10.1145\/3627673.3680276"},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"129447","DOI":"10.1016\/j.neucom.2025.129447","article-title":"H-vmunet: High-order vision mamba unet for medical image segmentation","volume":"23","author":"Wu","year":"2025","journal-title":"Neurocomputing"},{"key":"ref_26","doi-asserted-by":"crossref","first-page":"385","DOI":"10.3390\/e26050385","article-title":"Cascade residual multiscale convolution and mamba-structured unet for advanced brain tumor image segmentation","volume":"26","author":"Zhou","year":"2024","journal-title":"Entropy"},{"key":"ref_27","first-page":"81489","article-title":"Voxel mamba: Group-free state space models for point cloud based 3D object detection","volume":"37","author":"Zhang","year":"2024","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_28","first-page":"104092","article-title":"A local enhanced mamba network for hyperspectral image classification","volume":"133","author":"Wang","year":"2024","journal-title":"Int. J. Appl. Earth Obs. Geoinf."},{"key":"ref_29","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1109\/TGRS.2025.3592196","article-title":"Dualmamba: A lightweight spectral-spatial mamba-convolution network for hyperspectral image classification","volume":"63","author":"Sheng","year":"2024","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"ref_30","doi-asserted-by":"crossref","unstructured":"Ahmad, M., Butt, M.H.F., Khan, A.M., Mazzara, M., and Distefano, S. (2025). Spatial-spectral morphological mamba for hyperspectral image classification. Neurocomputing, 636.","DOI":"10.1016\/j.neucom.2025.129995"},{"key":"ref_31","doi-asserted-by":"crossref","unstructured":"Xie, D., Liang, L., Jin, L., Xu, J., and Li, M. (2015, January 9\u201312). SCUT-FBP: A benchmark dataset for facial beauty perception. Proceedings of the 2015 IEEE International Conference on Systems, Man, and Cybernetics, Hong Kong, China.","DOI":"10.1109\/SMC.2015.319"},{"key":"ref_32","doi-asserted-by":"crossref","unstructured":"Liang, L., Lin, L., Jin, L., Xie, D., and Li, M. (2018, January 20\u201324). SCUT-FBP5500: A diverse benchmark dataset for multi-paradigm facial beauty prediction. Proceedings of the 2018 24th International Conference on Pattern Recognition (ICPR), Beijing, China.","DOI":"10.1109\/ICPR.2018.8546038"},{"key":"ref_33","doi-asserted-by":"crossref","first-page":"2160002","DOI":"10.1142\/S0218001421600028","article-title":"Facial beauty prediction from facial parts using multi-task and multi-stream convolutional neural networks","volume":"35","author":"Vahdati","year":"2021","journal-title":"Int. J. Pattern Recognit. Artif. Intell."},{"key":"ref_34","doi-asserted-by":"crossref","first-page":"179","DOI":"10.3390\/electronics13010179","article-title":"Facial beauty prediction combined with multi-task learning of adaptive sharing policy and attentional feature fusion","volume":"13","author":"Gan","year":"2023","journal-title":"Electronics"},{"key":"ref_35","doi-asserted-by":"crossref","unstructured":"Vahdati, E., and Suen, C.Y. (2020, January 1\u20133). Facial beauty prediction using transfer and multi-task learning techniques. Proceedings of the International Conference on Pattern Recognition and Artificial Intelligence, Paris, France.","DOI":"10.1007\/978-3-030-59830-3_38"},{"key":"ref_36","doi-asserted-by":"crossref","first-page":"141627","DOI":"10.1109\/ACCESS.2019.2943604","article-title":"A comparison of loss weighting strategies for multi task learning in deep neural networks","volume":"7","author":"Gong","year":"2019","journal-title":"IEEE Access"},{"key":"ref_37","doi-asserted-by":"crossref","unstructured":"Liu, S., Johns, E., and Davison, A.J. (2019, January 15\u201320). End-to-end multi-task learning with attention. Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition, Long Beach, CA, USA.","DOI":"10.1109\/CVPR.2019.00197"},{"key":"ref_38","doi-asserted-by":"crossref","unstructured":"Chen, Z., Xie, L., Niu, J., Liu, X., and Wei, L. (2021, January 10\u201317). Visformer: The vision-friendly transformer. Proceedings of the IEEE\/CVF International Conference on Computer Vision (ICCV), Montreal, QC, Canada.","DOI":"10.1109\/ICCV48922.2021.00063"},{"key":"ref_39","doi-asserted-by":"crossref","unstructured":"Liu, Z., Lin, Y., Cao, Y., Hu, H., Wei, Y., Zhang, Z., Lin, S., and Guo, B. (2021, January 10\u201317). Swin transformer: Hierarchical vision transformer using shifted windows. Proceedings of the IEEE\/CVF International Conference on Computer Vision (ICCV), Montreal, QC, Canada.","DOI":"10.1109\/ICCV48922.2021.00986"},{"key":"ref_40","unstructured":"Woods, R.E., and Gonzalez, R.C. (2021). Digital Image Processing Third Edition, Electronic Industry Press."},{"key":"ref_41","first-page":"8583","article-title":"Scaling vision with sparse mixture of experts","volume":"34","author":"Riquelme","year":"2021","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_42","first-page":"269","article-title":"Tutel: Adaptive mixture-of-experts at scale","volume":"5","author":"Hwang","year":"2023","journal-title":"Proc. Mach. Learn. Syst."},{"key":"ref_43","doi-asserted-by":"crossref","unstructured":"Yun, S., Lee, H., Kim, J., and Shin, J. (2022, January 18\u201324). Patch-level representation learning for self-supervised vision transformers. Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), New Orleans, LA, USA.","DOI":"10.1109\/CVPR52688.2022.00817"},{"key":"ref_44","doi-asserted-by":"crossref","unstructured":"Zhang, Y., Cai, R., Chen, T., Zhang, G., Zhang, H., Chen, P.Y., Chang, S., Wang, Z., and Liu, S. (2023, January 1\u20136). Robust mixture-of-expert training for convolutional neural networks. Proceedings of the IEEE\/CVF International Conference on Computer Vision (ICCV), Paris, France.","DOI":"10.1109\/ICCV51070.2023.00015"},{"key":"ref_45","first-page":"7785","article-title":"Deep state space models for time series forecasting","volume":"31","author":"Rangapuram","year":"2018","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_46","first-page":"647","article-title":"Image transformers for classifying acute lymphoblastic leukemia","volume":"Volume 12033","author":"Cho","year":"2022","journal-title":"Medical Imaging 2022: Computer-Aided Diagnosis"},{"key":"ref_47","doi-asserted-by":"crossref","unstructured":"Yu, Z., Zhao, C., Wang, Z., Qin, Y., Su, Z., Li, X., Zhou, F., and Zhao, G. (2020, January 13\u201319). Searching central difference convolutional networks for face anti-spoofing. Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA.","DOI":"10.1109\/CVPR42600.2020.00534"},{"key":"ref_48","unstructured":"Dao, T., and Gu, A. (2024, January 21\u201327). Transformers are SSMs: Generalized models and efficient algorithms through structured state space duality. Proceedings of the 41st International Conference on Machine Learning (ICML), Honolulu, HI, USA."},{"key":"ref_49","unstructured":"Hu, E.J., Shen, Y., Wallis, P., Allen-Zhu, Z., Li, Y., Wang, S., Wang, L., and Chen, W. (2022, January 25\u201329). Lora: Low-rank adaptation of large language models. Proceedings of the International Conference on Learning Representations (ICLR), Virtual."},{"key":"ref_50","unstructured":"Rizzolatti, G., and Craighero, L. (2014). Spatial attention: Mechanisms and theories. Advances in Psychological Science, Psychology Press."},{"key":"ref_51","doi-asserted-by":"crossref","unstructured":"Wang, Q., Wu, B., Zhu, P., Li, P., Zuo, W., and Hu, Q. (2020, January 13\u201319). ECA-Net: Efficient channel attention for deep convolutional neural networks. Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA.","DOI":"10.1109\/CVPR42600.2020.01155"},{"key":"ref_52","doi-asserted-by":"crossref","unstructured":"Agiza, A., Neseem, M., and Reda, S. (2024, January 16\u201322). MTLoRA: Low-rank adaptation approach for efficient multi-task learning. Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Vancouver, BC, Canada.","DOI":"10.1109\/CVPR52733.2024.01533"},{"key":"ref_53","doi-asserted-by":"crossref","unstructured":"Ye, H., and Xu, D. (2022, January 23\u201327). Inverted pyramid multi-task transformer for dense scene understanding. Proceedings of the European Conference on Computer Vision(ECCV), Tel Aviv, Israel.","DOI":"10.1007\/978-3-031-19812-0_30"},{"key":"ref_54","unstructured":"Ye, H., and Xu, D. (2022, January 25\u201329). TaskPrompter: Spatial-channel multi-task prompting for dense scene understanding. Proceedings of the Eleventh International Conference on Learning Representations (ICLR), Virtual."},{"key":"ref_55","doi-asserted-by":"crossref","unstructured":"Ye, H., and Xu, D. (2024, January 16\u201322). DiffusionMTL: Learning multi-task denoising diffusion model from partially annotated data. Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Vancouver, BC, Canada.","DOI":"10.1109\/CVPR52733.2024.02641"},{"key":"ref_56","doi-asserted-by":"crossref","unstructured":"Yang, Y., Fu, H., Aviles-Rivero, A.I., Sch\u00f6nlieb, C.-B., and Zhu, L. (2023, January 8\u201312). Diffmic: Dual-guidance Diffusion Network for Medical Image Classification. Proceedings of the International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), Vancouver, BC, Canada.","DOI":"10.1007\/978-3-031-43987-2_10"},{"key":"ref_57","doi-asserted-by":"crossref","unstructured":"Shen, J., Wang, Y., and Luo, J. (2024). Cd-loop: A chromatin loop detection method based on the diffusion model. Front. Genet., 15.","DOI":"10.3389\/fgene.2024.1393406"},{"key":"ref_58","doi-asserted-by":"crossref","first-page":"13391","DOI":"10.1007\/s00500-022-07563-1","article-title":"Facial beauty prediction fusing transfer learning and broad learning system","volume":"27","author":"Gan","year":"2023","journal-title":"Soft Comput."},{"key":"ref_59","doi-asserted-by":"crossref","first-page":"5586","DOI":"10.1109\/TKDE.2021.3070203","article-title":"A Survey on Multi-Task Learning","volume":"34","author":"Zhang","year":"2022","journal-title":"IEEE Trans. Knowl. Data Eng."}],"container-title":["Symmetry"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2073-8994\/17\/10\/1600\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,9,26]],"date-time":"2025-09-26T08:47:10Z","timestamp":1758876430000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2073-8994\/17\/10\/1600"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,9,26]]},"references-count":59,"journal-issue":{"issue":"10","published-online":{"date-parts":[[2025,10]]}},"alternative-id":["sym17101600"],"URL":"https:\/\/doi.org\/10.3390\/sym17101600","relation":{},"ISSN":["2073-8994"],"issn-type":[{"value":"2073-8994","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,9,26]]}}}