{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,21]],"date-time":"2026-05-21T14:33:00Z","timestamp":1779373980259,"version":"3.53.1"},"reference-count":46,"publisher":"Springer Science and Business Media LLC","issue":"5","license":[{"start":{"date-parts":[[2022,3,16]],"date-time":"2022-03-16T00:00:00Z","timestamp":1647388800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2022,3,16]],"date-time":"2022-03-16T00:00:00Z","timestamp":1647388800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"DOI":"10.13039\/501100001809","name":"Natural Science Foundation of China","doi-asserted-by":"crossref","award":["61871325"],"award-info":[{"award-number":["61871325"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/501100001809","name":"Natural Science Foundation of China","doi-asserted-by":"crossref","award":["61671387"],"award-info":[{"award-number":["61671387"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"crossref"}]},{"name":"National Key Research and Development Program of China","award":["2018AAA0102803"],"award-info":[{"award-number":["2018AAA0102803"]}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Int J Comput Vis"],"published-print":{"date-parts":[[2022,5]]},"abstract":"<jats:title>Abstract<\/jats:title><jats:p>Although deep learning-based methods have dominated stereo matching leaderboards by yielding unprecedented disparity accuracy, their inference time is typically slow, <jats:italic>i.e.<\/jats:italic>, less than 4 FPS for a pair of 540p images. The main reason is that the leading methods employ time-consuming 3D convolutions applied to a 4D feature volume. A common way to speed up the computation is to downsample the feature volume, but this loses high-frequency details. To overcome these challenges, we propose a <jats:italic>displacement-invariant cost computation module<\/jats:italic> to compute the matching costs without needing a 4D feature volume. Rather, costs are computed by applying the same 2D convolution network on each disparity-shifted feature map pair independently. Unlike previous 2D convolution-based methods that simply perform context mapping between inputs and disparity maps, our proposed approach learns to match features between the two images. We also propose an entropy-based refinement strategy to refine the computed disparity map, which further improves the speed by avoiding the need to compute a second disparity map on the right image. Extensive experiments on standard datasets (SceneFlow, KITTI, ETH3D, and Middlebury) demonstrate that our method achieves competitive accuracy with much less inference time. On typical image sizes (<jats:italic>e.g.<\/jats:italic>, <jats:inline-formula><jats:alternatives><jats:tex-math>$$540\\times 960$$<\/jats:tex-math><mml:math xmlns:mml=\"http:\/\/www.w3.org\/1998\/Math\/MathML\">\n                  <mml:mrow>\n                    <mml:mn>540<\/mml:mn>\n                    <mml:mo>\u00d7<\/mml:mo>\n                    <mml:mn>960<\/mml:mn>\n                  <\/mml:mrow>\n                <\/mml:math><\/jats:alternatives><\/jats:inline-formula>), our method processes over 100 FPS on a desktop GPU, making our method suitable for time-critical applications such as autonomous driving. We also show that our approach generalizes well to unseen datasets, outperforming 4D-volumetric methods. We will release the source code to ensure the reproducibility.\n<\/jats:p>","DOI":"10.1007\/s11263-022-01595-8","type":"journal-article","created":{"date-parts":[[2022,3,16]],"date-time":"2022-03-16T06:02:36Z","timestamp":1647410556000},"page":"1196-1209","update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":8,"title":["Displacement-Invariant Cost Computation for Stereo Matching"],"prefix":"10.1007","volume":"130","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-1404-3610","authenticated-orcid":false,"given":"Yiran","family":"Zhong","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Charles","family":"Loop","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Wonmin","family":"Byeon","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Stan","family":"Birchfield","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yuchao","family":"Dai","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kaihao","family":"Zhang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Alexey","family":"Kamenev","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Thomas","family":"Breuel","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hongdong","family":"Li","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jan","family":"Kautz","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2022,3,16]]},"reference":[{"key":"1595_CR1","doi-asserted-by":"crossref","unstructured":"Badki, A., Gallo, O., Kautz, J., Sen, P. (2021). Binary TTC: A temporal geofence for autonomous navigation. In: CVPR.","DOI":"10.1109\/CVPR46437.2021.01275"},{"key":"1595_CR2","doi-asserted-by":"crossref","unstructured":"Badki, A., Troccoli, A., Kim, K., Kautz, J., Sen, P., Gallo, O. (2020). Bi3d: Stereo depth estimation via binary classifications. In: CVPR.","DOI":"10.1109\/CVPR42600.2020.00167"},{"key":"1595_CR3","doi-asserted-by":"crossref","unstructured":"Cai, C., Mordohai, P. (2020). Do end-to-end stereo algorithms under-utilize information? In: 3DV.","DOI":"10.1109\/3DV50981.2020.00047"},{"key":"1595_CR4","doi-asserted-by":"crossref","unstructured":"Cai, C., Poggi, M., Mattoccia, S., Mordohai, P. (2020). Matching-space stereo networks for cross-domain generalization. In: 3DV.","DOI":"10.1109\/3DV50981.2020.00046"},{"key":"1595_CR5","doi-asserted-by":"crossref","unstructured":"Chabra, R., Straub, J., Sweeney, C., Newcombe, R., Fuchs, H. (2019). StereoDRNet: Dilated residual StereoNet. In: CVPR.","DOI":"10.1109\/CVPR.2019.01206"},{"key":"1595_CR6","doi-asserted-by":"crossref","unstructured":"Chang, J.R., Chen, Y.S. (2018). Pyramid stereo matching network. In: CVPR.","DOI":"10.1109\/CVPR.2018.00567"},{"key":"1595_CR7","unstructured":"Cheng, X., Zhong, Y., Harandi, M., Dai, Y., Chang, X., Li, H., Drummond, T., Ge, Z. (2020). Hierarchical neural architecture search for deep stereo matching. In: NeurIPS."},{"key":"1595_CR8","doi-asserted-by":"crossref","unstructured":"Duggal, S., Wang, S., Ma, W.C., Hu, R., Urtasun, R. (2019). Deeppruner: Learning efficient stereo matching via differentiable patchmatch. In: ICCV.","DOI":"10.1109\/ICCV.2019.00448"},{"key":"1595_CR9","doi-asserted-by":"crossref","unstructured":"He, K., Zhang, X., Ren, S., Sun, J. (2014). Spatial pyramid pooling in deep convolutional networks for visual recognition. In: ECCV.","DOI":"10.1007\/978-3-319-10578-9_23"},{"key":"1595_CR10","doi-asserted-by":"crossref","unstructured":"Henkel, R.D. (1997). Constructing the cyclopean view. In: ICANN.","DOI":"10.1007\/BFb0020268"},{"key":"1595_CR11","doi-asserted-by":"publisher","first-page":"143","DOI":"10.1016\/j.procs.2016.05.305","volume":"80","author":"D Hernandez-Juarez","year":"2016","unstructured":"Hernandez-Juarez, D., Chac\u00f3n, A., Espinosa, A., V\u00e1zquez, D., Moure, J., & L\u00f3pez, A. (2016). Embedded real-time stereo estimation via semi-global matching on the GPU. Procedia Computer Science, 80, 143\u2013153.","journal-title":"Procedia Computer Science"},{"key":"1595_CR12","doi-asserted-by":"crossref","unstructured":"Hirschmuller, H. (2008). Stereo processing by semiglobal matching and mutual information. Intell: IEEE Trans. Pattern Anal. Mach.","DOI":"10.1109\/TPAMI.2007.1166"},{"key":"1595_CR13","doi-asserted-by":"crossref","unstructured":"Ilg, E., Saikia, T., Keuper, M., Brox, T. (2018). Occlusions, motion and depth boundaries with a generic network for disparity, optical flow or scene flow estimation. In: ECCV.","DOI":"10.1007\/978-3-030-01258-8_38"},{"key":"1595_CR14","unstructured":"Janai, J., G\u00fcney, F., Behl, A., Geiger, A. (2017). Computer vision for autonomous vehicles: Problems, datasets and state-of-the-art. Arxiv."},{"key":"1595_CR15","volume-title":"Foundations of Cyclopean Perception","author":"B Julesz","year":"1971","unstructured":"Julesz, B. (1971). Foundations of Cyclopean Perception. Chicago: The University of Chicago Press."},{"key":"1595_CR16","doi-asserted-by":"crossref","unstructured":"Kendall, A., Martirosyan, H., Dasgupta, S., Henry, P., Kennedy, R., Bachrach, A., Bry, A. (2017). End-to-end learning of geometry and context for deep stereo regression. In: ICCV.","DOI":"10.1109\/ICCV.2017.17"},{"key":"1595_CR17","doi-asserted-by":"crossref","unstructured":"Khamis, S., Fanello, S., Rhemann, C., Kowdle, A., Valentin, J., Izadi, S. (2018). StereoNet: Guided hierarchical refinement for real-time edge-aware depth prediction. In: ECCV.","DOI":"10.1007\/978-3-030-01267-0_35"},{"key":"1595_CR18","doi-asserted-by":"crossref","unstructured":"Kim, S., Kim, S., Min, D., Sohn, K. (2019). Laf-net: Locally adaptive fusion networks for stereo confidence estimation. In: CVPR.","DOI":"10.1109\/CVPR.2019.00029"},{"key":"1595_CR19","doi-asserted-by":"crossref","unstructured":"Kuzmin, A., Mikushin, D., Lempitsky, V. (2017). End-to-end learning of cost-volume aggregation for real-time dense stereo. In: MLSP.","DOI":"10.1109\/MLSP.2017.8168183"},{"key":"1595_CR20","doi-asserted-by":"crossref","unstructured":"Lee, H., Shin, Y. (2019). Real-time stereo matching network with high accuracy. In: ICIP.","DOI":"10.1109\/ICIP.2019.8803514"},{"key":"1595_CR21","doi-asserted-by":"crossref","unstructured":"Liang, Z., Feng, Y., Guo, Y., Liu, H., Chen, W., Qiao, L., Zhou, L., Zhang, J. (2018). Learning for disparity estimation through feature constancy. In: CVPR.","DOI":"10.1109\/CVPR.2018.00297"},{"key":"1595_CR22","doi-asserted-by":"crossref","unstructured":"Mayer, N., Ilg, E., H\u00e4usser, P., Fischer, P., Cremers, D., Dosovitskiy, A., Brox, T. (2016). A large dataset to train convolutional networks for disparity, optical flow, and scene flow estimation. In: CVPR.","DOI":"10.1109\/CVPR.2016.438"},{"key":"1595_CR23","doi-asserted-by":"crossref","unstructured":"Menze, M., Geiger, A. (2015). Object scene flow for autonomous vehicles. In: CVPR.","DOI":"10.1109\/CVPR.2015.7298925"},{"key":"1595_CR24","doi-asserted-by":"crossref","unstructured":"Pang, J., Sun, W., Ren, J.S., Yang, C., Yan, Q. (2017). Cascade residual learning: A two-stage convolutional neural network for stereo matching. In: ICCVW.","DOI":"10.1109\/ICCVW.2017.108"},{"key":"1595_CR25","doi-asserted-by":"crossref","unstructured":"Poggi, M., Aleotti, F., Tosi, F., Mattoccia, S. (2020). On the uncertainty of self-supervised monocular depth estimation. In: CVPR.","DOI":"10.1109\/CVPR42600.2020.00329"},{"key":"1595_CR26","doi-asserted-by":"crossref","unstructured":"Ronneberger, O., Fischer, P., Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. In: Medical Image Computing and Computer-Assisted Intervention.","DOI":"10.1007\/978-3-319-24574-4_28"},{"key":"1595_CR27","doi-asserted-by":"crossref","unstructured":"Scharstein, D., Hirschm\u00fcller, H., Kitajima, Y., Krathwohl, G., Ne\u0161i\u0107, N., Wang, X., Westling, P. (2014). High-resolution stereo datasets with subpixel-accurate ground truth. In: Pattern Recognition.","DOI":"10.1007\/978-3-319-11752-2_3"},{"key":"1595_CR28","unstructured":"Scharstein, D., Szeliski, R. (1998). Stereo matching with nonlinear diffusion. In: IJCV."},{"key":"1595_CR29","unstructured":"Scharstein, D., Szeliski, R. (2002). A taxonomy and evaluation of dense two-frame stereo correspondence algorithms. In: IJCV."},{"key":"1595_CR30","doi-asserted-by":"crossref","unstructured":"Sch\u00f6ps, T., Sch\u00f6nberger, J.L., Galliani, S., Sattler, T., Schindler, K., Pollefeys, M., Geiger, A. (2017). A multi-view stereo benchmark with high-resolution images and multi-camera videos. In: CVPR.","DOI":"10.1109\/CVPR.2017.272"},{"key":"1595_CR31","doi-asserted-by":"crossref","unstructured":"Seki, A., Pollefeys, M. (2017). SGM-Nets: Semi-global matching with neural networks. In: CVPR.","DOI":"10.1109\/CVPR.2017.703"},{"key":"1595_CR32","doi-asserted-by":"crossref","unstructured":"Smolyanskiy, N., Kamenev, A., Birchfield, S. (2018). On the importance of stereo for accurate depth estimation: An efficient semi-supervised deep neural network approach. In: CVPRW.","DOI":"10.1109\/CVPRW.2018.00147"},{"key":"1595_CR33","doi-asserted-by":"crossref","unstructured":"Taniai, T., Matsushita, Y., Sato, Y., Naemura, T. (2018). Continuous 3D Label Stereo Matching using Local Expansion Moves. In: TPAMI.","DOI":"10.1109\/TPAMI.2017.2766072"},{"key":"1595_CR34","doi-asserted-by":"crossref","unstructured":"Teed, Z., Deng, J. (2020). Raft: Recurrent all-pairs field transforms for optical flow. In: ECCV.","DOI":"10.24963\/ijcai.2021\/662"},{"key":"1595_CR35","doi-asserted-by":"crossref","unstructured":"Tonioni, A., Tosi, F., Poggi, M., Mattoccia, S., Di\u00a0Stefano, L. (2019). Real-time self-adaptive deep stereo. In: CVPR.","DOI":"10.1109\/CVPR.2019.00028"},{"key":"1595_CR36","doi-asserted-by":"crossref","unstructured":"Weber, M., Humenberger, M., Kubinger, W. (2009). A very fast census-based stereo matching implementation on a graphics processing unit. In: ICCVW.","DOI":"10.1109\/ICCVW.2009.5457622"},{"key":"1595_CR37","doi-asserted-by":"crossref","unstructured":"Yang, G., Manela, J., Happold, M., Ramanan, D. (2019). Hierarchical deep stereo matching on high-resolution images. In: CVPR.","DOI":"10.1109\/CVPR.2019.00566"},{"key":"1595_CR38","doi-asserted-by":"crossref","unstructured":"Yao, Y., Luo, Z., Li, S., Shen, T., Fang, T., Quan, L. (2019). Recurrent mvsnet for high-resolution multi-view stereo depth inference. In: CVPR.","DOI":"10.1109\/CVPR.2019.00567"},{"key":"1595_CR39","doi-asserted-by":"crossref","unstructured":"Yin, Z., Darrell, T., Yu, F. (2019). Hierarchical discrete distribution decomposition for match density estimation. In: CVPR.","DOI":"10.1109\/CVPR.2019.00620"},{"issue":"1","key":"1595_CR40","first-page":"2287","volume":"17","author":"J \u017dbontar","year":"2016","unstructured":"\u017dbontar, J., & LeCun, Y. (2016). Stereo matching by training a convolutional neural network to compare image patches. J Mach Learn Res, 17(1), 2287\u20132318.","journal-title":"J Mach Learn Res"},{"key":"1595_CR41","doi-asserted-by":"crossref","unstructured":"Zhang, C., Li, Z., Cheng, Y., Cai, R., Chao, H., Rui, Y. (2015). Meshstereo: A global stereo model with mesh alignment regularization for view interpolation. In: ICCV.","DOI":"10.1109\/ICCV.2015.238"},{"key":"1595_CR42","doi-asserted-by":"crossref","unstructured":"Zhang, F., Prisacariu, V., Yang, R., Torr, P.H. (2019). GA-Net: Guided aggregation net for end-to-end stereo matching. In: CVPR.","DOI":"10.1109\/CVPR.2019.00027"},{"key":"1595_CR43","doi-asserted-by":"crossref","unstructured":"Zhang, F., Qi, X., Yang, R., Prisacariu, V., Wah, B., Torr, P.H.S. (2020). Domain-invariant stereo matching networks. In: ECCV.","DOI":"10.1007\/978-3-030-58536-5_25"},{"key":"1595_CR44","unstructured":"Zhong, Y., Dai, Y., Li, H. (2017). Self-supervised learning for stereo matching with self-improving ability. In: arXiv:1709.00930."},{"key":"1595_CR45","doi-asserted-by":"crossref","unstructured":"Zhong, Y., Li, H., Dai, Y. (2018). Open-world stereo video matching with deep RNN. In: ECCV.","DOI":"10.1007\/978-3-030-01216-8_7"},{"key":"1595_CR46","doi-asserted-by":"crossref","unstructured":"Zhou, H., Ummenhofer, B., Brox, T. (2018). Deeptam: Deep tracking and mapping. In: ECCV.","DOI":"10.1007\/978-3-030-01270-0_50"}],"container-title":["International Journal of Computer Vision"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s11263-022-01595-8.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s11263-022-01595-8\/fulltext.html","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s11263-022-01595-8.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,4,27]],"date-time":"2022-04-27T11:14:22Z","timestamp":1651058062000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s11263-022-01595-8"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,3,16]]},"references-count":46,"journal-issue":{"issue":"5","published-print":{"date-parts":[[2022,5]]}},"alternative-id":["1595"],"URL":"https:\/\/doi.org\/10.1007\/s11263-022-01595-8","relation":{},"ISSN":["0920-5691","1573-1405"],"issn-type":[{"value":"0920-5691","type":"print"},{"value":"1573-1405","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,3,16]]},"assertion":[{"value":"18 February 2021","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"14 February 2022","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"16 March 2022","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}}]}}