{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,11,3]],"date-time":"2026-11-03T05:47:44Z","timestamp":1793684864957,"version":"4.5.1"},"reference-count":46,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2025,12,12]],"date-time":"2025-12-12T00:00:00Z","timestamp":1765497600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,1,15]],"date-time":"2026-01-15T00:00:00Z","timestamp":1768435200000},"content-version":"vor","delay-in-days":34,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Sci Rep"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Object detection in aerial imagery, particularly from UAVs and drones, presents unique challenges such as small and extremely small-scale object distributions. To address these drawbacks, the present research proposes an efficient YOLOv12 model. The first core contribution of our work is the introduction of the Area-Attention C2f (A2C2F) module, which fuses multi-head MLP blocks with localized area-attention mechanisms to strengthen spatial feature learning while maintaining lightweight computation. Second, the model integrates the Cross Stage Partial with Kernel size 2 (C3K2) module, a lightweight variant of the C3 block, which reduces convolutional complexity and computational cost while maintaining feature extraction capability for efficient detection. Third, a multi-scale fusion strategy by stacking A2C2F blocks along with Concat and Upsample operations ensures the preservation of high-resolution features for detecting small-scale objects, a common challenge in drone-captured data. Fourth, a novel detection head with decoupled classification and regression branches, enhanced by attention-guided feature fusion, improves prediction consistency and boosts inference speed. Extensive experiments conducted on the VisDrone dataset demonstrate that the proposed framework effectively detects objects with dimensions smaller than 3\u20135 pixels, where traditional YOLO variants typically fail, proving its robustness in extremely small-scale detection. The proposed model achieves a Precision of 69.1%, Recall of 48.5%, and F1-score of 56.99%, outperforming recent models. Notably, it achieves a mAP@50 of 58.8% and mAP@0.5:0.95 of 40.9%, showing substantial improvement over prior methods. Despite utilizing 59.1\u00a0M parameters and 198.6 GFLOPs, the model maintains an efficient inference rate of approximately 40 FPS on the A100 GPU, confirming its suitability for real-time UAV edge deployment. When compared to two-stage and anchor-based models, our model delivers higher AP scores at a lower input resolution of 640\u2009\u00d7\u2009640 pixels, making it a highly effective and computationally efficient framework for UAV-based small and extremely small object detection.<\/jats:p>","DOI":"10.1038\/s41598-025-31803-7","type":"journal-article","created":{"date-parts":[[2025,12,12]],"date-time":"2025-12-12T07:25:06Z","timestamp":1765524306000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":14,"title":["An efficient YOLOv12-based framework for detecting extremely small-scale objects"],"prefix":"10.1038","volume":"16","author":[{"given":"A.","family":"Chandrashekhar","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"B.","family":"Satyanarayana","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Rajani Reddy","family":"Gorrepati","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ponduri","family":"Vasanthi","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kothala Lakshmi","family":"Prasanna","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2025,12,12]]},"reference":[{"issue":"19","key":"31803_CR1","doi-asserted-by":"publisher","first-page":"21839","DOI":"10.1109\/JSEN.2021.3103612","volume":"21","author":"H Luo","year":"2021","unstructured":"Luo, H., Wang, P., Chen, H. & Xu, M. Object detection method based on shallow feature fusion and semantic information enhancement. IEEE Sens. J. 21 (19), 21839\u201321851 (2021).","journal-title":"IEEE Sens. J."},{"issue":"5","key":"31803_CR2","doi-asserted-by":"publisher","first-page":"823","DOI":"10.1109\/TLA.2022.9693567","volume":"20","author":"Z Song","year":"2022","unstructured":"Song, Z., Zhang, Y., Liu, Y., Yang, K. & Sun, M. Feature fusion-based detection for small objects. IEEE Lat. Am. Trans. 20 (5), 823\u2013830 (2022).","journal-title":"IEEE Lat. Am. Trans."},{"key":"31803_CR3","doi-asserted-by":"publisher","first-page":"56214","DOI":"10.1109\/ACCESS.2021.3072067","volume":"9","author":"R Jin","year":"2021","unstructured":"Jin, R., Lv, J., Li, B., Ye, J. & Lin, D. Toward efficient object detection in aerial images using extreme-scale metric learning. IEEE Access. 9, 56214\u201356227 (2021).","journal-title":"IEEE Access."},{"issue":"1","key":"31803_CR4","doi-asserted-by":"publisher","first-page":"361","DOI":"10.1007\/s00500-021-06407-8","volume":"26","author":"W Zhan","year":"2022","unstructured":"Zhan, W. et al. An improved Yolov5 real-time detection method for small objects captured by UAV. Soft. Comput. 26 (1), 361\u2013373 (2022).","journal-title":"Soft. Comput."},{"key":"31803_CR5","unstructured":"Bochkovskiy, A. C.-Y.Wang, and H.-Y.M. Liao,\u2018\u2018YOLOv4:Opti-malspeedand accuracy of object detection,\u2019\u2019 arXiv:2004.10934. [Online]. Available: http:\/\/arxiv.org\/abs\/2004.10934\u00a0(2020)."},{"key":"31803_CR6","unstructured":"Ultralytics. YOLOv5.Accessed:Oct.30,2020.[Online]. (2020). Available: https:\/\/github.com\/ultralytics\/yolov5.git"},{"key":"31803_CR7","doi-asserted-by":"crossref","unstructured":"Hei Law and Jia Deng. Cornernet: Detecting objects as paired keypoints. In ProceedingsoftheEuropeanconferenceoncomputervision(ECCV). : 734\u2013750. (2018).","DOI":"10.1007\/978-3-030-01264-9_45"},{"key":"31803_CR8","unstructured":"Li, Z. et al. Light-headr-cnn:Indefenseoftwo-stageobjectdetector. arXivpreprintarXiv:1711.07264, (2017)."},{"key":"31803_CR9","unstructured":"Lin, T. Y. et al. Featurepyramidnetworksforobjectdetection.InProceedingsoftheIEEEconferenceon computer vision and pattern recognition. (2117\u20132125). (2017)."},{"key":"31803_CR10","doi-asserted-by":"crossref","unstructured":"Wan, J., Zhang, B., Zhao, Y., Du, Y. & Tong, Z. VistrongerDet:StrongerVisual Information for Object Detection in VisDrone Images. In Proceedings of the IEEE\/CVF International Conference on Computer Vision. (2820\u20132829). (2021).","DOI":"10.1109\/ICCVW54120.2021.00316"},{"issue":"12","key":"31803_CR11","doi-asserted-by":"publisher","first-page":"36491","DOI":"10.1007\/s11042-023-15773-4","volume":"83","author":"P Vasanthi","year":"2024","unstructured":"Vasanthi, P. Multi-Head-Self-Attention based YOLOv5X-transformer for multi-scale object detection. Multimedia Tools Appl. 83 (12), 36491\u201336517 (2024).","journal-title":"Multimedia Tools Appl."},{"key":"31803_CR12","doi-asserted-by":"publisher","first-page":"809","DOI":"10.1016\/j.neunet.2023.06.020","volume":"165","author":"P Vasanthi","year":"2023","unstructured":"Vasanthi, P. & Mohan, L. A reliable anchor regenerative-based transformer model for x-small and dense objects recognition. Neural Netw. 165, 809\u2013829 (2023).","journal-title":"Neural Netw."},{"key":"31803_CR13","doi-asserted-by":"publisher","first-page":"104378","DOI":"10.1016\/j.bspc.2022.104378","volume":"80","author":"L Kothala","year":"2023","unstructured":"Kothala, L., Prasanna, P., Jonnala & Sitaramanjaneya Reddy, G. Localization of mixed intracranial hemorrhages by using a ghost convolution-based YOLO network. Biomed. Signal Process. Control. 80, 104378 (2023).","journal-title":"Biomed. Signal Process. Control"},{"key":"31803_CR14","doi-asserted-by":"crossref","unstructured":"Kothala, L. P. & Sitaramanjaneya Reddy Guntur TL-LFF net: transfer learning based lighter, faster, and frozen network for the detection of multi-scale mixed intracranial hemorrhages through genetic optimization algorithm. Int. J. Mach. Learn. Cybernet.\u00a01\u201327. (2024).","DOI":"10.1007\/s13042-024-02324-y"},{"key":"31803_CR15","doi-asserted-by":"crossref","unstructured":"Yue, M., Zhang, L., Huang, J. & Zhang, H. Lightweight and efficient tiny-object detection based on improved YOLOv8n for UAV aerial images. Drones.\u00a087. (2024).","DOI":"10.3390\/drones8070276"},{"key":"31803_CR16","first-page":"1","volume":"73","author":"L Jiang","year":"2024","unstructured":"Jiang, L. et al. Mffsodnet: multiscale feature fusion small object detection network for uav aerial images. IEEE Trans. Instrum. Meas. 73, 1\u201314 (2024).","journal-title":"IEEE Trans. Instrum. Meas."},{"key":"31803_CR17","doi-asserted-by":"crossref","unstructured":"Jia Su, Z. J. Y. & Qin, Liang, B. MPE-YOLO: enhanced small target detection in aerial imaging. Sci. Rep.\u00a014. (2024).","DOI":"10.1038\/s41598-024-68934-2"},{"key":"31803_CR18","doi-asserted-by":"crossref","unstructured":"Tang, L., Yun, L., Chen, Z. & Cheng, F. HRYNet: A highly robust Yolo network for complex road traffic object detection. Sensors.\u00a024 (2), (2024).","DOI":"10.3390\/s24020642"},{"key":"31803_CR19","doi-asserted-by":"crossref","unstructured":"Yanli Shi, Y. J. & Zhang, X. FocusDet: an efficient object detector for small object. Scientific Reports. 14. pmid:38730236. (2024).","DOI":"10.1038\/s41598-024-61136-w"},{"key":"31803_CR20","doi-asserted-by":"crossref","unstructured":"Minze Li, T. Z. Y. & Chen, Huang, W. TA-YOLO: a lightweight small object detection model based on multi-dimensional trans-attention module for remote sensing images. Complex. Intell. Syst.\u00a010. (2024).","DOI":"10.1007\/s40747-024-01448-6"},{"issue":"4","key":"31803_CR21","doi-asserted-by":"publisher","first-page":"304","DOI":"10.1109\/TLA.2024.10472961","volume":"22","author":"P Zhang","year":"2024","unstructured":"Zhang, P., Zhang, G., Yang, K. & APNet Accurate positioning deformable Convolution for uav image object detection. IEEE Lat. Am. Trans. 22 (4), 304\u2013311 (2024).","journal-title":"IEEE Lat. Am. Trans."},{"key":"31803_CR22","doi-asserted-by":"crossref","unstructured":"Wu, W. et al. EUAVDet: an efficient and lightweight object detector for uav aerial images with an edge-based computing platform. Drones.\u00a086. (2024).","DOI":"10.3390\/drones8060261"},{"key":"31803_CR23","doi-asserted-by":"crossref","unstructured":"Zhou, L. et al. MFEFNet: A multi-scale feature information extraction and fusion network for multi-scale object detection in UAV aerial images. Drones. 85. (2024).","DOI":"10.3390\/drones8050186"},{"key":"31803_CR24","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/TGRS.2024.3510383","volume":"62","author":"S Huang","year":"2024","unstructured":"Huang, S., Lin, C., Jiang, X. & Qu, Z. Brstd: Bio-inspired remote sensing tiny object detection. IEEE Trans. Geosci. Remote Sens. 62, 1\u201315 (2024).","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"31803_CR25","doi-asserted-by":"crossref","unstructured":"Li, Y. et al. Sod-yolo: Small-object-detection algorithm based on improved yolov8 for uav images, Remote Sensing, 16, [Online]. Available: https:\/\/www.mdpi.com\/2072-4292\/16\/16\/3057\u00a0(2024).","DOI":"10.3390\/rs16163057"},{"key":"31803_CR26","doi-asserted-by":"crossref","unstructured":"Wang, S., Jiang, H., Yang, J., Ma, X. & Chen, J. Amfef-detr: An end-to-end adaptive multi-scale feature extraction and fusion object detection network based on uav aerial images.\u00a0Drones. 8\u00a0(10). [Online]. Available: https:\/\/www.mdpi.com\/2504-446X\/8\/10\/523\u00a0(2024).","DOI":"10.3390\/drones8100523"},{"key":"31803_CR27","doi-asserted-by":"crossref","unstructured":"Tan, S., Duan, Z. & Pu, L. Multi-scale object detection in uav images based on adaptive feature fusion.\u00a0PLos One, 19\u00a0(3), 1\u201321. [Online]. Available: https:\/\/doi.org\/10.1371\/journal.pone.0300120\u00a0(2024).","DOI":"10.1371\/journal.pone.0300120"},{"key":"31803_CR28","doi-asserted-by":"crossref","unstructured":"Jing, S., Lv, H., Zhao, Y. & Liu, H. and Sun M., Mvt: Multi-vision transformer for event-based small target detection.\u00a0Remote Sensing, 16\u00a0(9), [Online]. Available: https:\/\/www.mdpi.com\/2072-4292\/16\/9\/1641\u00a0(2024).","DOI":"10.3390\/rs16091641"},{"key":"31803_CR29","doi-asserted-by":"crossref","unstructured":"Zhang, H., Sun, W., Sun, C., He, R. & Zhang, Y. HSP-YOLOv8: UAV aerial photography small target detection algorithm. Drones. 89. (2024).","DOI":"10.3390\/drones8090453"},{"key":"31803_CR30","doi-asserted-by":"publisher","first-page":"104682","DOI":"10.1016\/j.dsp.2024.104682","volume":"154","author":"Vasanthi","year":"2024","unstructured":"Vasanthi & Mohan, L. Efficient YOLOv8 algorithm for extreme small-scale object detection. Digit. Signal Proc. 154, 104682 (2024).","journal-title":"Digit. Signal Proc."},{"issue":"1","key":"31803_CR31","doi-asserted-by":"publisher","first-page":"015287","DOI":"10.1088\/2631-8695\/adc223","volume":"7","author":"V Ponduri","year":"2025","unstructured":"Ponduri, V. et al. Highly efficient YOLOV9 model for detecting extremely small-scale objects. Eng. Res. Express. 7 (1), 015287 (2025).","journal-title":"Eng. Res. Express"},{"issue":"12","key":"31803_CR32","doi-asserted-by":"publisher","first-page":"e0315267","DOI":"10.1371\/journal.pone.0315267","volume":"19","author":"D Tang","year":"2024","unstructured":"Tang, D., Tang, S., Fan, Z., LCFF-Net & A lightweight cross-scale feature fusion network for tiny target detection in UAV aerial imagery. PloS One. 19 (12), e0315267 (2024).","journal-title":"PloS One"},{"key":"31803_CR33","doi-asserted-by":"crossref","unstructured":"Zhao wei Cai and Nuno Vasconcelos.Cascader-cnn:Delvingintohigh-qualityobject detection. In Proceedings of the IEEE conference on computer vision and pattern recognition, 6154\u20136162. (2018).","DOI":"10.1109\/CVPR.2018.00644"},{"key":"31803_CR34","doi-asserted-by":"publisher","first-page":"24330","DOI":"10.1109\/TITS.2022.3203715","volume":"23","author":"X Li","year":"2022","unstructured":"Li, X., Jiang, Z., Chen, Y. & Wang, X. An anchor-free lightweight deep convolutional network for vehicle detection in aerial images. IEEE Trans. Intell. Transp. Syst. 23, 12,24330\u201324342. https:\/\/doi.org\/10.1109\/TITS.2022.3203715 (2022).","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"key":"31803_CR35","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/TIM.2023.3346488","volume":"5008516","author":"Y Wang","year":"2024","unstructured":"Wang, Y., Liu, X., Zhang, Q. & Zhao, J. An instrument indication acquisition algorithm based on lightweight deep convolutional neural network and hybrid attention fine-grained features. IEEE Trans. Instrum. Meas. 73 Art. 5008516, 1\u201316. https:\/\/doi.org\/10.1109\/TIM.2023.3346488 (2024).","journal-title":"IEEE Trans. Instrum. Meas. 73 Art"},{"key":"31803_CR36","doi-asserted-by":"publisher","DOI":"10.1038\/s40494-025-01565-6","author":"T Zhang","year":"2025","unstructured":"Zhang, T., Liu, H. & Zhao, W. An algorithm based on lightweight semantic features for ancient mural element object detection. Npj Herit. Sci. https:\/\/doi.org\/10.1038\/s40494-025-01565-6 (2025)., 13, art. 70.","journal-title":"Npj Herit. Sci."},{"key":"31803_CR37","doi-asserted-by":"crossref","unstructured":"Chen, F., Zhang, H. & Li, Y. Finger vein recognition algorithm based on lightweight deep convolutional neural network. IEEE Transactions on Instrumentation and Measurement. 71, Art. 5000413, 1\u201313, https:\/\/doi.org\/10.1109\/TIM.2021.3132332\u00a0(2022).","DOI":"10.1109\/TIM.2021.3132332"},{"key":"31803_CR38","first-page":"4716","volume":"29","author":"J Zhao","year":"2020","unstructured":"Zhao, J., Liu, J., Fan, D., Zhang, Y. & Cheng, M. M. Dynamic feature integration for simultaneous detection of salient object, edge and skeleton. IEEE Trans. Image Process. 29, 4716\u20134729 (2020).","journal-title":"IEEE Trans. Image Process."},{"key":"31803_CR39","first-page":"6257","volume":"30","author":"Y Luo","year":"2021","unstructured":"Luo, Y., Zhao, H., Li, Z. & Bai, H. Rethinking the U-shape structure for salient object detection. IEEE Trans. Image Process. 30, 6257\u20136269 (2021).","journal-title":"IEEE Trans. Image Process."},{"key":"31803_CR40","unstructured":"Sun, X., Chen, L. & Wang, Z. Diffusion model-based path follower for a salamander-like robot. IEEE Trans. Neural Networks Learn. Syst.\u00a036 (5), 1\u201312 (2025)."},{"key":"31803_CR41","unstructured":"Chen, J., Yang, R. & Li, Q. Towards efficient salient object detection via U-shape architecture search. Knowl. Based Syst.\u00a0295, 111986, (2025)."},{"key":"31803_CR42","doi-asserted-by":"publisher","DOI":"10.1109\/TGRS.2023.3305728","author":"X Liu","year":"2023","unstructured":"Liu, X., Zhang, Z. & Wang, L. SmallTrack: wavelet pooling and graph enhanced classification for UAV small object tracking. IEEE Trans. Geosci. Remote Sens. https:\/\/doi.org\/10.1109\/TGRS.2023.3305728 (2023).","journal-title":"IEEE Trans. Geosci. Remote Sens."},{"key":"31803_CR43","doi-asserted-by":"publisher","unstructured":"Wang, W. et al. FMTrack: Frequency-aware interaction and multi-expert fusion for RGB-T tracking. IEEE Trans. Circuits Syst. Video Technol. 35 (1). https:\/\/doi.org\/10.1109\/TCSVT.2025.3601598 (2025).","DOI":"10.1109\/TCSVT.2025.3601598"},{"key":"31803_CR44","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/TCSVT.2025.3549953","volume":"35","author":"H Li","year":"2025","unstructured":"Li, H. et al. AVLTrack: dynamic sparse learning for aerial vision-language tracking. IEEE Trans. Circuits Syst. Video Technol. 35, 3,1\u201310. https:\/\/doi.org\/10.1109\/TCSVT.2025.3549953 (2025).","journal-title":"IEEE Trans. Circuits Syst. Video Technol."},{"key":"31803_CR45","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/TIM.2024.3440378","volume":"74","author":"Y Zhang","year":"2024","unstructured":"Zhang, Y. et al. Handling occlusion in uav visual tracking with query-guided redemption. IEEE Trans. Instrum. Meas. 74, 1\u20139. https:\/\/doi.org\/10.1109\/TIM.2024.3440378 (2024).","journal-title":"IEEE Trans. Instrum. Meas."},{"key":"31803_CR46","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/TITS.2025.3581391","volume":"26","author":"S Chen","year":"2025","unstructured":"Chen, S. et al. Target-distractor aware UAV tracking via global agent. IEEE Trans. Intell. Transp. Syst. 26, 4,1\u201311. https:\/\/doi.org\/10.1109\/TITS.2025.3581391 (2025).","journal-title":"IEEE Trans. Intell. Transp. Syst."}],"container-title":["Scientific Reports"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.nature.com\/articles\/s41598-025-31803-7.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41598-025-31803-7","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41598-025-31803-7.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,1,15]],"date-time":"2026-01-15T12:34:40Z","timestamp":1768480480000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.nature.com\/articles\/s41598-025-31803-7"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,12,12]]},"references-count":46,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2026,12]]}},"alternative-id":["31803"],"URL":"https:\/\/doi.org\/10.1038\/s41598-025-31803-7","relation":{},"ISSN":["2045-2322"],"issn-type":[{"value":"2045-2322","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,12,12]]},"assertion":[{"value":"8 October 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"5 December 2025","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"12 December 2025","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"The authors declare no competing interests.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interests"}}],"article-number":"2062"}}