{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,26]],"date-time":"2026-05-26T10:04:56Z","timestamp":1779789896885,"version":"3.53.1"},"reference-count":47,"publisher":"Springer Science and Business Media LLC","issue":"6","license":[{"start":{"date-parts":[[2026,5,26]],"date-time":"2026-05-26T00:00:00Z","timestamp":1779753600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,5,26]],"date-time":"2026-05-26T00:00:00Z","timestamp":1779753600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"DOI":"10.13039\/100020595","name":"National Science and Technology Council","doi-asserted-by":"publisher","award":["NSTC 113-2634-F-A49-007"],"award-info":[{"award-number":["NSTC 113-2634-F-A49-007"]}],"id":[{"id":"10.13039\/100020595","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Int J Comput Vis"],"published-print":{"date-parts":[[2026,6]]},"abstract":"Abstract<\/jats:title>\n \n An ultralightweight self-supervised stereo matching network, called US\n \n \n $$^3$$<\/jats:tex-math>\n \n \n \n \n 3<\/mml:mn>\n <\/mml:mmultiscripts>\n <\/mml:math>\n <\/jats:alternatives>\n <\/jats:inline-formula>\n Net, which requires only 12K parameters, is designed for efficient and accurate depth estimation using resource-constrained devices. US\n \n \n $$^3$$<\/jats:tex-math>\n \n \n \n \n 3<\/mml:mn>\n <\/mml:mmultiscripts>\n <\/mml:math>\n <\/jats:alternatives>\n <\/jats:inline-formula>\n Net incorporates two key innovations: a low-complexity feature extraction module and a soft occlusion detection approach for performance improvement. First, we design a low-complexity feature extraction module to reduce the computational burden while preserving structural details necessary for stereo matching. By refining the encoder backbone and aggregation module, our design ensures a better balance between model complexity and accuracy. Second, to address occlusion-related errors in disparity estimation, we propose a novel occlusion detection method, called Depth-Aware Geometric Soft Occlusion (DAGSO), to adaptively define the occlusion confidence scores based on depth information. DAGSO can effectively mitigate false occlusions in distant regions and can improve the accuracy of disparity estimation. Experimental results using KITTI datasets demonstrate that US\n \n \n $$^3$$<\/jats:tex-math>\n \n \n \n \n 3<\/mml:mn>\n <\/mml:mmultiscripts>\n <\/mml:math>\n <\/jats:alternatives>\n <\/jats:inline-formula>\n Net achieves state-of-the-art performance in terms of model complexity and depth estimation accuracy. It outperforms previous self-supervised stereo matching methods and monocular depth estimation methods in metrics such as AbsRel, SqRel, RMSE, and RMSElog at a reduction of parameter size by 47% compared with ES\n \n \n $$^3$$<\/jats:tex-math>\n \n \n \n \n 3<\/mml:mn>\n <\/mml:mmultiscripts>\n <\/mml:math>\n <\/jats:alternatives>\n <\/jats:inline-formula>\n Net (23K). This makes US\n \n \n $$^3$$<\/jats:tex-math>\n \n \n \n \n 3<\/mml:mn>\n <\/mml:mmultiscripts>\n <\/mml:math>\n <\/jats:alternatives>\n <\/jats:inline-formula>\n Net a practical solution for real-time depth estimation on edge devices such as drones and autonomous systems. Code is available at:\n https:\/\/github.com\/g830319ag\/US3Net.<\/jats:ext-link>\n <\/jats:p>","DOI":"10.1007\/s11263-026-02868-2","type":"journal-article","created":{"date-parts":[[2026,5,26]],"date-time":"2026-05-26T09:35:19Z","timestamp":1779788119000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["US$$^3$$Net: Ultralightweight Self-Supervised Stereo Matching Network using Depth-Aware Geometric Soft Occlusion"],"prefix":"10.1007","volume":"134","author":[{"given":"Po-Chung","family":"Jen","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tzu-Chi","family":"Liu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"I-Sheng","family":"Fang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hsiao-Chieh","family":"Wen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chia-Lun","family":"Hsu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ping-Yang","family":"Chen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chang-Hsing","family":"Lee","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5581-850X","authenticated-orcid":false,"given":"Yong-Sheng","family":"Chen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,5,26]]},"reference":[{"key":"2868_CR1","doi-asserted-by":"crossref","unstructured":"Chang, J.R., & Chen, Y.S. (2018). Pyramid stereo matching network. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp 5410\u20135418","DOI":"10.1109\/CVPR.2018.00567"},{"key":"2868_CR2","doi-asserted-by":"crossref","unstructured":"Chang, J.R., Chang, P.C.,& Chen, Y.S. (2020). Attention-aware feature aggregation for real-time stereo matching on edge devices. In: Proceedings of the Asian Conference on Computer Vision (ACCV)","DOI":"10.1007\/978-3-030-69525-5_22"},{"key":"2868_CR3","doi-asserted-by":"crossref","unstructured":"Cheng, J., Yin, W., Wang, K., Chen, X., Wang, S., & Yang, X. (2024). Adaptive fusion of single-view and multi-view depth for autonomous driving. In: Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition, pp 10138\u201310147","DOI":"10.1109\/CVPR52733.2024.00966"},{"key":"2868_CR4","doi-asserted-by":"crossref","unstructured":"Chi, C., Wang, Q., Hao, T., Guo, P., & Yang, X. (2021). Feature-level collaboration: Joint unsupervised learning of optical flow, stereo depth and camera motion. In: Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition, pp 2463\u20132473","DOI":"10.1109\/CVPR46437.2021.00249"},{"key":"2868_CR5","doi-asserted-by":"crossref","unstructured":"Chung, J., Oh, J.,& Lee, K.M. (2024). Depth-regularized optimization for 3d gaussian splatting in few-shot images. In: Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition, pp 811\u2013820","DOI":"10.1109\/CVPRW63382.2024.00086"},{"key":"2868_CR6","doi-asserted-by":"crossref","unstructured":"Cordts, M., Omran, M., Ramos, S., Rehfeld, T., Enzweiler, M., Benenson, R., & Schiele, B. (2016). The cityscapes dataset for semantic urban scene understanding. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 3213\u20133223","DOI":"10.1109\/CVPR.2016.350"},{"key":"2868_CR7","doi-asserted-by":"crossref","unstructured":"Deng, J., Dong, W., Socher, R., Li, L. J., Li, K., & Fei-Fei, L. (2009). Imagenet: A large-scale hierarchical image database. In: 2009 IEEE conference on computer vision and pattern recognition, Ieee, pp 248\u2013255","DOI":"10.1109\/CVPR.2009.5206848"},{"key":"2868_CR8","unstructured":"Eigen, D., Puhrsch, C.,& Fergus, R. (2014). Depth map prediction from a single image using a multi-scale deep network. Advances in neural information processing systems 27"},{"key":"2868_CR9","doi-asserted-by":"crossref","unstructured":"Fang, I., Wen, H.C., Hsu, C.L., Jen, P. C., Chen, P. Y., & Chen, Y. S. (2023). Es3net: Accurate and efficient edge-based self-supervised stereo matching network. In: Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition, pp 4472\u20134481","DOI":"10.1109\/CVPRW59228.2023.00470"},{"key":"2868_CR10","doi-asserted-by":"crossref","unstructured":"Geiger, A., Lenz, P.,& Urtasun, R. (2012). Are we ready for autonomous driving? the kitti vision benchmark suite. In: 2012 IEEE conference on computer vision and pattern recognition, IEEE, pp 3354\u20133361","DOI":"10.1109\/CVPR.2012.6248074"},{"key":"2868_CR11","doi-asserted-by":"crossref","unstructured":"Geiger, A., Lenz, P., Stiller, C., & Urtasun, R. (2013). Vision meets robotics: The kitti dataset. International Journal of Robotics Research (IJRR)","DOI":"10.1177\/0278364913491297"},{"key":"2868_CR12","doi-asserted-by":"crossref","unstructured":"Godard, C., Mac\u00a0Aodha, O.,& Brostow, G.J. (2017). Unsupervised monocular depth estimation with left-right consistency. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 270\u2013279","DOI":"10.1109\/CVPR.2017.699"},{"key":"2868_CR13","doi-asserted-by":"crossref","unstructured":"Godard, C., Mac\u00a0Aodha, O.,& Firman, M., et\u00a0al. (2019). Digging into self-supervised monocular depth estimation. In: Proceedings of the IEEE\/CVF international conference on computer vision, pp 3828\u20133838","DOI":"10.1109\/ICCV.2019.00393"},{"key":"2868_CR14","doi-asserted-by":"crossref","unstructured":"Haase, D.,& Amthor, M. (2020). Rethinking depthwise separable convolutions: How intra-kernel correlations lead to improved mobilenets. In: Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition, pp 14600\u201314609","DOI":"10.1109\/CVPR42600.2020.01461"},{"key":"2868_CR15","unstructured":"Howard, A.G., Zhu, M., Chen, B., Kalenichenko, D., Wang, W., Weyand, T., & Adam, H. (2017). Mobilenets: Efficient convolutional neural networks for mobile vision applications. arXiv preprint arXiv:1704.04861"},{"key":"2868_CR16","unstructured":"Kingma, D.P. (2014). Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980"},{"key":"2868_CR17","doi-asserted-by":"crossref","unstructured":"Lai, H.Y., Tsai, Y.H.,& Chiu, W.C. (2019). Bridging stereo matching and optical flow via spatiotemporal correspondence. In: Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition, pp 1890\u20131899","DOI":"10.1109\/CVPR.2019.00199"},{"key":"2868_CR18","doi-asserted-by":"crossref","unstructured":"Li, A.,& Yuan, Z. (2018). Occlusion aware stereo matching via cooperative unsupervised learning. In: Asian Conference on Computer Vision, Springer, pp 197\u2013213","DOI":"10.1007\/978-3-030-20876-9_13"},{"issue":"7","key":"2868_CR19","doi-asserted-by":"publisher","first-page":"7457","DOI":"10.1109\/TITS.2021.3070403","volume":"23","author":"A Li","year":"2021","unstructured":"Li, A., Yuan, Z., Ling, Y., et al. (2021). Unsupervised occlusion-aware stereo matching with directed disparity smoothing. IEEE Transactions on Intelligent Transportation Systems, 23(7), 7457\u20137468.","journal-title":"IEEE Transactions on Intelligent Transportation Systems"},{"key":"2868_CR20","doi-asserted-by":"crossref","unstructured":"Menze, M.,& Geiger, A. (2015). Object scene flow for autonomous vehicles. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 3061\u20133070","DOI":"10.1109\/CVPR.2015.7298925"},{"key":"2868_CR21","unstructured":"Paszke, A., Gross, S., Massa, F., Lerer, A., Bradbury, J., Chanan, G., & Chintala, S. (2019). Pytorch: An imperative style, high-performance deep learning library. Advances in neural information processing systems 32"},{"key":"2868_CR22","unstructured":"Peng, L., Deng, D.,& Cai, D. (2020). Geometry-based occlusion-aware unsupervised stereo matching for autonomous driving. arXiv preprint arXiv:2010.10700"},{"key":"2868_CR23","doi-asserted-by":"crossref","unstructured":"Piccinelli, L., Yang, Y.H., Sakaridis, C., Segu, M., Li, S., Van Gool, L., & Yu, F. (2024). Unidepth: Universal monocular metric depth estimation. In: Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition, pp 10106\u201310116","DOI":"10.1109\/CVPR52733.2024.00963"},{"key":"2868_CR24","doi-asserted-by":"crossref","unstructured":"Pjetri, A., Caprasecca, S., Taccari, L., Simoncini, M., Monteagudo, H. P., Walter, W., & Bagdanov, A. D. (2024). A new dataset for monocular depth estimation under viewpoint shifts. arXiv preprint arXiv:2409.17851","DOI":"10.1109\/IV64158.2025.11097590"},{"key":"2868_CR25","doi-asserted-by":"crossref","unstructured":"Poggi, M., Aleotti, F., Tosi, F., & Mattoccia, S. (2018a). Towards real-time unsupervised monocular depth estimation on cpu. In: 2018 IEEE\/RSJ international conference on intelligent robots and systems (IROS), IEEE, pp 5848\u20135854","DOI":"10.1109\/IROS.2018.8593814"},{"key":"2868_CR26","doi-asserted-by":"crossref","unstructured":"Poggi, M., Tosi, F., & Mattoccia, S. (2018b). Learning monocular depth estimation with unsupervised trinocular assumptions. In: 2018 International conference on 3d vision (3DV), IEEE, pp 324\u2013333","DOI":"10.1109\/3DV.2018.00045"},{"issue":"10","key":"2868_CR27","doi-asserted-by":"publisher","first-page":"17342","DOI":"10.1109\/TITS.2022.3157265","volume":"23","author":"M Poggi","year":"2022","unstructured":"Poggi, M., Tosi, F., Aleotti, F., et al. (2022). Real-time self-supervised monocular depth estimation without gpu. IEEE Transactions on Intelligent Transportation Systems, 23(10), 17342\u201317353.","journal-title":"IEEE Transactions on Intelligent Transportation Systems"},{"key":"2868_CR28","doi-asserted-by":"crossref","unstructured":"Shi, W., Caballero, J., Husz\u00e1r, F., Totz, J., Aitken, A. P., Bishop, R., & Wang, Z. (2016). Real-time single image and video super-resolution using an efficient sub-pixel convolutional neural network. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 1874\u20131883","DOI":"10.1109\/CVPR.2016.207"},{"key":"2868_CR29","doi-asserted-by":"crossref","unstructured":"Shi, Y., Singh, M.K., Cai, H., & Porikli, F. (2024). Decotr: Enhancing depth completion with 2D and 3D attentions. In: Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition, pp 10736\u201310746","DOI":"10.1109\/CVPR52733.2024.01021"},{"key":"2868_CR30","doi-asserted-by":"crossref","unstructured":"Tosi, F., Aleotti, F., Poggi, M., & Mattoccia, S. (2019). Learning monocular depth estimation infusing traditional stereo knowledge. In: Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition, pp 9799\u20139809","DOI":"10.1109\/CVPR.2019.01003"},{"key":"2868_CR31","doi-asserted-by":"crossref","unstructured":"Wang, Y., Yang, Y., Yang, Z., Zhao, L., Wang, P., & Xu, W. (2018a). Occlusion aware unsupervised learning of optical flow. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 4884\u20134893","DOI":"10.1109\/CVPR.2018.00513"},{"key":"2868_CR32","unstructured":"Wang, Y., Yang, Z., Wang, P., Yang, Y., Luo, C., & Xu, W. (2018b). Joint unsupervised learning of optical flow and depth by watching stereo videos. arXiv preprint arXiv:1810.03654"},{"key":"2868_CR33","doi-asserted-by":"crossref","unstructured":"Wang, Y., Lai, Z., Huang, G., Wang, B.H., Van Der Maaten, L., Campbell, M., & Weinberger, K. Q. (2019a). Anytime stereo image depth estimation on mobile devices. In: 2019 international conference on robotics and automation (ICRA), IEEE, pp 5893\u20135900","DOI":"10.1109\/ICRA.2019.8794003"},{"key":"2868_CR34","doi-asserted-by":"crossref","unstructured":"Wang, Y., Wang, P., Yang, Z., Luo, C., Yang, Y., & Xu, W. (2019b). Unos: Unified unsupervised optical-flow and stereo-depth estimation by watching videos. In: Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition, pp 8071\u20138081","DOI":"10.1109\/CVPR.2019.00826"},{"issue":"4","key":"2868_CR35","doi-asserted-by":"publisher","first-page":"600","DOI":"10.1109\/TIP.2003.819861","volume":"13","author":"Z Wang","year":"2004","unstructured":"Wang, Z., Bovik, A. C., Sheikh, H. R., et al. (2004). Image quality assessment: from error visibility to structural similarity. IEEE transactions on image processing, 13(4), 600\u2013612.","journal-title":"IEEE transactions on image processing"},{"key":"2868_CR36","doi-asserted-by":"crossref","unstructured":"Watson, J., Firman, M., Brostow, G.J. & Turmukhambetov, D. (2019). Self-supervised monocular depth hints. In: Proceedings of the IEEE\/CVF international conference on computer vision, pp 2162\u20132171","DOI":"10.1109\/ICCV.2019.00225"},{"key":"2868_CR37","doi-asserted-by":"crossref","unstructured":"Wolf, Y., Bracha, A., Kimmel, R. (2024). Gs2mesh: Surface reconstruction from gaussian splatting via novel stereo views. In: European Conference on Computer Vision, Springer, pp 207\u2013224","DOI":"10.1007\/978-3-031-73024-5_13"},{"key":"2868_CR38","doi-asserted-by":"crossref","unstructured":"Xu, G. Wang, X., Ding, X., & Yang, X. (2023). Iterative geometry encoding volume for stereo matching. In: Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition, pp 21919\u201321928","DOI":"10.1109\/CVPR52729.2023.02099"},{"key":"2868_CR39","doi-asserted-by":"crossref","unstructured":"Xu, H., Zhang, J. (2020). Aanet: Adaptive aggregation network for efficient stereo matching. In: Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition, pp 1959\u20131968","DOI":"10.1109\/CVPR42600.2020.00203"},{"key":"2868_CR40","doi-asserted-by":"crossref","unstructured":"Yang, G., Song, X., Huang, C., Deng, Z., Shi, J., & Zhou, B. (2019). Drivingstereo: A large-scale dataset for stereo matching in autonomous driving scenarios. In: Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition, pp 899\u2013908","DOI":"10.1109\/CVPR.2019.00099"},{"key":"2868_CR41","unstructured":"Yang, R., Li, X., Bai, J., & Du, J. (2024). Pseudo-stereo inputs: A solution to the occlusion challenge in self-supervised stereo matching. arXiv preprint arXiv:2410.02534"},{"key":"2868_CR42","doi-asserted-by":"crossref","unstructured":"Yang, Z., Wang, P., Wang, Y., Xu, W., & Nevatia, R. (2018). Every pixel counts: Unsupervised geometry learning with holistic 3d motion understanding. In: Proceedings of the European conference on computer vision (ECCV) workshops, pp 0\u20130","DOI":"10.1007\/978-3-030-11021-5_43"},{"key":"2868_CR43","doi-asserted-by":"crossref","unstructured":"Zhang, F., Prisacariu, V., Yang, R., & Torr, P.H.S. (2019). Ga-net: Guided aggregation net for end-to-end stereo matching. In: Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition, pp 185\u2013194","DOI":"10.1109\/CVPR.2019.00027"},{"key":"2868_CR44","doi-asserted-by":"crossref","unstructured":"Zhang, X., Xue, Y., Jia, S., & Pei, X. (2024a). Ccdepth: A lightweight self-supervised depth estimation network with enhanced interpretability. arXiv preprint arXiv:2409.19933","DOI":"10.1109\/ITSC58415.2024.10919195"},{"issue":"21","key":"2868_CR45","doi-asserted-by":"publisher","first-page":"7013","DOI":"10.3390\/s24217013","volume":"24","author":"Y Zhang","year":"2024","unstructured":"Zhang, Y., Chu, L., Wang, Z., et al. (2024). A novel panorama depth estimation framework for autonomous driving scenarios based on a vision transformer. Sensors, 24(21), 7013.","journal-title":"Sensors"},{"key":"2868_CR46","unstructured":"Zhong, Y., Dai, Y., Li, H. (2017). Self-supervised learning for stereo matching with self-improving ability. arXiv preprint arXiv:1709.00930"},{"key":"2868_CR47","doi-asserted-by":"crossref","unstructured":"Zhong, Y., Li, H., Dai, Y. (2018). Open-world stereo video matching with deep rnn. In: Proceedings of the European Conference on Computer Vision (ECCV), pp 101\u2013116","DOI":"10.1007\/978-3-030-01216-8_7"}],"container-title":["International Journal of Computer Vision"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s11263-026-02868-2.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s11263-026-02868-2","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s11263-026-02868-2.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,26]],"date-time":"2026-05-26T09:35:34Z","timestamp":1779788134000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s11263-026-02868-2"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,5,26]]},"references-count":47,"journal-issue":{"issue":"6","published-print":{"date-parts":[[2026,6]]}},"alternative-id":["2868"],"URL":"https:\/\/doi.org\/10.1007\/s11263-026-02868-2","relation":{},"ISSN":["0920-5691","1573-1405"],"issn-type":[{"value":"0920-5691","type":"print"},{"value":"1573-1405","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,5,26]]},"assertion":[{"value":"15 March 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"27 April 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"26 May 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}}],"article-number":"286"}}