{"url":"/sota/depth-completion-on-kitti-depth-completion","task":{"name":"Depth Completion","url":"/task/depth-completion","note":null},"dataset":{"name":"KITTI Depth Completion","url":null},"category":"Computer Vision","categories":["Computer Vision"],"category_note":null,"description":"The **Depth Completion** task is a sub-problem of depth estimation. In the sparse-to-dense depth completion problem, one wants to infer the dense depth map of a 3-D scene given an RGB image and its corresponding sparse reconstruction in the form of a sparse depth map obtained either from computational methods such as SfM (Strcuture-from-Motion) or active sensors such as lidar or structured light sensors. \r\n\r\n<span class=\"description-source\">Source: [LiStereo: Generate Dense Depth Maps from LIDAR and Stereo Imagery ](https://arxiv.org/abs/1905.02744)</span>,\r\n<span class=\"description-source\">[Unsupervised Depth Completion from Visual Inertial Odometry](https://arxiv.org/abs/1905.08616)</span>","description_from":"task","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","rank":"the archive's row order at snapshot; not re-ranked","rows_end_at":"2025-07-28","rows_withheld_as_spam":0,"metric_values":"the archive's strings, untouched"},"metrics":["RMSE","MAE","iRMSE","iMAE","Runtime [ms]"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"RMSE":"lower","MAE":"lower","iRMSE":null,"iMAE":null,"Runtime [ms]":null}},"counts":{"rows":16,"rows_with_code":11,"rows_with_paper_page":16,"rows_dated":16,"rows_using_additional_data":0},"rows":[{"rank_in_archive_order":1,"model":"SemAttNet","metrics":{"RMSE":"709.41"},"uses_additional_data":false,"paper_date":"2022-04-28","paper":"/paper/semattnet-towards-attention-based-semantic","paper_url":"https://arxiv.org/abs/2204.13635v1","paper_title":"SemAttNet: Towards Attention-based Semantic Aware Guided Depth Completion","code":"https://github.com/danishnazir/SemAttNet","n_code_links":1,"syntology":null},{"rank_in_archive_order":2,"model":"RigNet","metrics":{"RMSE":"713.44"},"uses_additional_data":false,"paper_date":"2021-07-29","paper":"/paper/rignet-repetitive-image-guided-network-for","paper_url":"https://arxiv.org/abs/2107.13802v5","paper_title":"RigNet: Repetitive Image Guided Network for Depth Completion","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":3,"model":"PENet","metrics":{"MAE":"210.55","RMSE":"730.08","Runtime [ms]":"32","iMAE":"0.94","iRMSE":"2.17"},"uses_additional_data":false,"paper_date":"2021-03-01","paper":"/paper/towards-precise-and-efficient-image-guided","paper_url":"https://arxiv.org/abs/2103.00783v3","paper_title":"PENet: Towards Precise and Efficient Image Guided Depth Completion","code":"https://github.com/JUGGHM/PENet_ICRA2021","n_code_links":3,"syntology":null},{"rank_in_archive_order":4,"model":"NLSPN","metrics":{"MAE":"199.59","RMSE":"741.68","Runtime [ms]":"220","iMAE":"0.84","iRMSE":"1.99"},"uses_additional_data":false,"paper_date":"2020-07-20","paper":"/paper/non-local-spatial-propagation-network-for","paper_url":"https://arxiv.org/abs/2007.10042v1","paper_title":"Non-Local Spatial Propagation Network for Depth Completion","code":"https://github.com/zzangjinsun/NLSPN_ECCV20","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":4,"n_samples":5,"n_pointer_only_licence":0}},{"rank_in_archive_order":5,"model":"FusionNet (RGB_guide&certainty)","metrics":{"MAE":"215.02","RMSE":"772.87","Runtime [ms]":"20","iMAE":"0.93","iRMSE":"2.19"},"uses_additional_data":false,"paper_date":"2019-02-14","paper":"/paper/sparse-and-noisy-lidar-completion-with-rgb","paper_url":"http://arxiv.org/abs/1902.05356v1","paper_title":"Sparse and noisy LiDAR completion with RGB guidance and uncertainty","code":"https://github.com/wvangansbeke/Sparse-Depth-Completion","n_code_links":1,"syntology":null},{"rank_in_archive_order":6,"model":"DVMN","metrics":{"MAE":"220.37","RMSE":"776.31","iMAE":"0.94","iRMSE":"2.21"},"uses_additional_data":false,"paper_date":"2021-07-14","paper":"/paper/dvmn-dense-validity-mask-network-for-depth","paper_url":"https://arxiv.org/abs/2107.06709v1","paper_title":"DVMN: Dense Validity Mask Network for Depth Completion","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":7,"model":"NConv-CNN-L2","metrics":{"MAE":"233","RMSE":"830","Runtime [ms]":"20"},"uses_additional_data":false,"paper_date":"2018-11-05","paper":"/paper/confidence-propagation-through-cnns-for","paper_url":"https://arxiv.org/abs/1811.01791v2","paper_title":"Confidence Propagation through CNNs for Guided Sparse Depth Regression","code":"https://github.com/abdo-eldesokey/nconv","n_code_links":1,"syntology":null},{"rank_in_archive_order":8,"model":"NConv-CNN-L1","metrics":{"MAE":"208","RMSE":"859","Runtime [ms]":"20"},"uses_additional_data":false,"paper_date":"2018-11-05","paper":"/paper/confidence-propagation-through-cnns-for","paper_url":"https://arxiv.org/abs/1811.01791v2","paper_title":"Confidence Propagation through CNNs for Guided Sparse Depth Regression","code":"https://github.com/abdo-eldesokey/nconv","n_code_links":1,"syntology":null},{"rank_in_archive_order":9,"model":"Spade-RGBsD","metrics":{"MAE":"235","RMSE":"918","Runtime [ms]":"70"},"uses_additional_data":false,"paper_date":"2018-08-02","paper":"/paper/sparse-and-dense-data-with-cnns-depth","paper_url":"http://arxiv.org/abs/1808.00769v2","paper_title":"Sparse and Dense Data with CNNs: Depth Completion and Semantic Segmentation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":10,"model":"HMS-Net","metrics":{"MAE":"258","RMSE":"937"},"uses_additional_data":false,"paper_date":"2018-08-27","paper":"/paper/hms-net-hierarchical-multi-scale-sparsity","paper_url":"https://arxiv.org/abs/1808.08685v2","paper_title":"HMS-Net: Hierarchical Multi-scale Sparsity-invariant Network for Sparse Depth Completion","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":11,"model":"Spade-sD","metrics":{"MAE":"248","RMSE":"1035","Runtime [ms]":"40"},"uses_additional_data":false,"paper_date":"2018-08-02","paper":"/paper/sparse-and-dense-data-with-cnns-depth","paper_url":"http://arxiv.org/abs/1808.00769v2","paper_title":"Sparse and Dense Data with CNNs: Depth Completion and Semantic Segmentation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":12,"model":"KBNet","metrics":{"MAE":"256.76","RMSE":"1069.47","Runtime [ms]":"16","iMAE":"1.02","iRMSE":"2.95"},"uses_additional_data":false,"paper_date":"2021-08-24","paper":"/paper/unsupervised-depth-completion-with-calibrated","paper_url":"https://arxiv.org/abs/2108.10531v2","paper_title":"Unsupervised Depth Completion with Calibrated Backprojection Layers","code":"https://github.com/alexklwong/calibrated-backprojection-network","n_code_links":1,"syntology":null},{"rank_in_archive_order":13,"model":"ScaffNet-FusionNet","metrics":{"MAE":"280.76","RMSE":"1121.93","iMAE":"1.15","iRMSE":"3.30"},"uses_additional_data":false,"paper_date":"2021-06-06","paper":"/paper/learning-topology-from-synthetic-data-for","paper_url":"https://arxiv.org/abs/2106.02994v3","paper_title":"Learning Topology from Synthetic Data for Unsupervised Depth Completion","code":"https://github.com/alexklwong/learning-topology-synthetic-data","n_code_links":1,"syntology":null},{"rank_in_archive_order":14,"model":"VOICED","metrics":{"MAE":"299.41","RMSE":"1169.97","Runtime [ms]":"20","iMAE":"1.20","iRMSE":"3.56"},"uses_additional_data":false,"paper_date":"2019-05-15","paper":"/paper/190508616","paper_url":"https://arxiv.org/abs/1905.08616v4","paper_title":"Unsupervised Depth Completion from Visual Inertial Odometry","code":"https://github.com/alexklwong/unsupervised-depth-completion-visual-inertial-odometry","n_code_links":2,"syntology":null},{"rank_in_archive_order":15,"model":"NConv-CNN","metrics":{"MAE":"360","RMSE":"1268","Runtime [ms]":"10"},"uses_additional_data":false,"paper_date":"2018-11-05","paper":"/paper/confidence-propagation-through-cnns-for","paper_url":"https://arxiv.org/abs/1811.01791v2","paper_title":"Confidence Propagation through CNNs for Guided Sparse Depth Regression","code":"https://github.com/abdo-eldesokey/nconv","n_code_links":1,"syntology":null},{"rank_in_archive_order":16,"model":"SparseConvs","metrics":{"MAE":"481","RMSE":"1601","Runtime [ms]":"10"},"uses_additional_data":false,"paper_date":"2017-08-22","paper":"/paper/sparsity-invariant-cnns","paper_url":"http://arxiv.org/abs/1708.06500v2","paper_title":"Sparsity Invariant CNNs","code":"https://github.com/PeterTor/sparse_convolution","n_code_links":1,"syntology":null}],"since_archive":{"claim":"Results that newer papers report for their own method, placed here by Syntology. A model pointed at the cell in the paper's own table; the number was read from that cell and checked against this leaderboard's metric, dataset, split and scale; an independent check that saw this leaderboard's other rows and every other leaderboard on the same dataset accepted it. Not reviewed by the paper's authors or by the archive's editors, and not ranked against the archive rows.","extraction_file_present":true,"measurement":{"test_papers":883,"papers_with_output":881,"judged_true":108,"judged":110,"wilson95_lower":0.9361,"measured_on":"2026-09-24","frozen_commit":"0e3de0df94"},"measurement_note":"blind adjudication of accepted entries on a held-out split of archive papers, rules frozen before the test","coverage":{"sentence":"Syntology has checked 6,795 of the 9,581 papers on this site that are newer than the archive; results from the others appear after they are checked.","complete":false,"papers_newer_than_archive":9581,"papers_checked":6795,"papers_extracted_not_yet_verified":0,"boards_without_verdict":2,"papers_not_yet_extracted":2785},"order":"newest first by month (arXiv date, else the arXiv-id month), then arXiv id descending","columns":[],"entries":[]},"syntology":{"read_at":"2026-09-24T18:15:14+00:00","claim":"Per row: N of M harvested code samples from that row's paper executed on a synthesized fixture; the other M-N are unverified. Not a reproduction of the row's number; not a correctness claim. n_pointer_only_licence counts samples the site points at rather than redistributes (a licence axis, independent of ran/unverified).","rows_with_graph_line":1,"rows_with_any_sample_ran":1,"distinct_papers_with_graph_line":1,"distinct_papers_with_any_sample_ran":1,"samples_over_distinct_papers":{"n_ran":1,"n_unverified":4,"n_samples":5,"n_pointer_only_licence":0,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":1,"n_unverified":4,"n_samples":5,"n_pointer_only_licence":0,"note":"row-weighted: a paper behind several rows is counted once per row; inflated relative to samples_over_distinct_papers by design, kept for readers summing the per-row syntology blocks"}}}