{"url":"/dataset/cmu-panoptic","name":"Panoptic","full_name":"CMU Panoptic Studio","description_markdown":"**CMU Panoptic** is a large scale dataset providing 3D pose annotations (1.5 millions) for multiple people engaging social activities. It contains 65 videos (5.5 hours) with multi-view annotations, but only 17 of them are in multi-person scenario and have the camera parameters.\r\n\t\r\n**Massively Multiview System**\r\n\r\n*  480 VGA camera views\r\n*  30+ HD views\r\n*  10 RGB-D sensors\r\n*  Hardware-based sync\r\n*  Calibration\r\n* Interesting Scenes with Labels\r\n\r\n**Multiple people**\r\n\r\n* Socially interacting groups\r\n* 3D body pose\r\n* 3D facial landmarks\r\n* Transcripts + speaker ID\r\n\r\n**Hardware setup**\r\n\r\n* 480 VGA cameras, 640 x 480 resolution, 25 fps, synchronized among themselves using a hardware clock\r\n* 31 HD cameras, 1920 x 1080 resolution, 30 fps, synchronized among themselves using a hardware clock, timing aligned with VGA cameras\r\n* 10 Kinect Ⅱ Sensors. 1920 x 1080 (RGB), 512 x 424 (depth), 30 fps, timing aligned among themselves and other sensors\r\n5 DLP Projectors. synchronized with HD cameras\r\n\r\nSource: [Single-Stage Multi-Person Pose Machines](https://arxiv.org/abs/1908.09220)\r\nImage Source: [http://domedb.perception.cs.cmu.edu/](http://domedb.perception.cs.cmu.edu/)","description_withheld":null,"homepage":"http://domedb.perception.cs.cmu.edu/","introduced_date":"2015-01-01","introduced_date_note":null,"introduced_by":{"paper":"/paper/panoptic-studio-a-massively-multiview-system-1","title":"Panoptic Studio: A Massively Multiview System for Social Motion Capture","first_author":"Hanbyul Joo","url":null},"license":{"name":"Custom (non-commercial)","url":"http://domedb.perception.cs.cmu.edu/#:~:License"},"modalities":[{"name":"Images","url":"/datasets/modality/images"}],"tasks":[{"name":"3D Human Pose Estimation","url":"/task/3d-human-pose-estimation","datasets_with_task":"/datasets/task/3d-human-pose-estimation"},{"name":"3D Human Pose Tracking","url":"/task/3d-human-pose-tracking","datasets_with_task":"/datasets/task/3d-human-pose-tracking"},{"name":"Head Pose Estimation","url":"/task/head-pose-estimation","datasets_with_task":"/datasets/task/head-pose-estimation"},{"name":"3D Multi-Person Pose Estimation","url":"/task/3d-multi-person-pose-estimation","datasets_with_task":"/datasets/task/3d-multi-person-pose-estimation"}],"languages":[],"variants":["Panoptic"],"data_loaders":[{"repo":"https://github.com/open-mmlab/mmpose","url":"https://github.com/open-mmlab/mmpose/blob/master/docs/tasks/2d_hand_keypoint.md#cmu-panoptic-handdb","frameworks":["pytorch"]},{"repo":"https://gitlab.com/Percipiote/skelda","url":"https://gitlab.com/Percipiote/skelda","frameworks":[]}],"num_papers_in_archive":131,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[{"leaderboard":"/sota/3d-multi-person-pose-estimation-on-cmu","task":"3D Multi-Person Pose Estimation","dataset_variant":"Panoptic","rows":20,"metrics":["Average MPJPE (mm)"],"first_row_in_archive_order":{"model":"TesseTrack","paper":"/paper/tessetrack-end-to-end-learnable-multi-person","metrics":{"Average MPJPE (mm)":"7.3"},"code_links":[]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/3d-human-pose-estimation-on-cmu-panoptic","task":"3D Human Pose Estimation","dataset_variant":"Panoptic","rows":9,"metrics":["Average MPJPE (mm)"],"first_row_in_archive_order":{"model":"TesseTrack Multi-View (5 views)","paper":"/paper/tessetrack-end-to-end-learnable-multi-person","metrics":{"Average MPJPE (mm)":"7.3"},"code_links":[]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/head-pose-estimation-on-panoptic","task":"Head Pose Estimation","dataset_variant":"Panoptic","rows":6,"metrics":["Geodesic Error (GE)"],"first_row_in_archive_order":{"model":"WRHP-6D-Opal","paper":"/paper/on-the-representation-and-methodology-for-1","metrics":{"Geodesic Error (GE)":"7.45"},"code_links":[{"title":"pcr-upm/opal23_headpose","url":"https://github.com/pcr-upm/opal23_headpose"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/3d-human-pose-tracking-on-cmu-panoptic","task":"3D Human Pose Tracking","dataset_variant":"Panoptic","rows":1,"metrics":["3DMOTA"],"first_row_in_archive_order":{"model":"TesseTrack","paper":"/paper/tessetrack-end-to-end-learnable-multi-person","metrics":{"3DMOTA":"94.1"},"code_links":[]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"}],"papers_with_a_benchmark_row":[{"paper":"/paper/rapidposetriangulation-multi-view-multi","title":"RapidPoseTriangulation: Multi-view Multi-person Whole-body Human Pose Triangulation in a Millisecond","date":"2025-03-27","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/on-the-representation-and-methodology-for-1","title":"On the representation and methodology for wide and short range head pose estimation","date":"2024-01-11","rows_on_this_dataset":4,"code_links":1,"syntology":null},{"paper":"/paper/instance-aware-contrastive-learning-for","title":"Instance-aware Contrastive Learning for Occluded Human Mesh Reconstruction","date":"2024-01-01","rows_on_this_dataset":2,"code_links":1,"syntology":null},{"paper":"/paper/ivt-an-end-to-end-instance-guided-video","title":"IVT: An End-to-End Instance-guided Video Transformer for 3D Pose Estimation","date":"2022-08-06","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/quickpose-real-time-multi-view-multi-person","title":"QuickPose: Real-time Multi-view Multi-person Pose Estimation in Crowded Scenes","date":"2022-07-22","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/faster-voxelpose-real-time-3d-human-pose","title":"Faster VoxelPose: Real-time 3D Human Pose Estimation by Orthographic Projection","date":"2022-07-22","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":4,"samples_ran":0,"samples_unverified":4,"pointer_only_for_licence":4,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/dynamic-graph-reasoning-for-multi-person-3d","title":"Dynamic Graph Reasoning for Multi-person 3D Pose Estimation","date":"2022-07-22","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/virtualpose-learning-generalizable-3d-human","title":"VirtualPose: Learning Generalizable 3D Human Pose Models from Virtual Data","date":"2022-07-20","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/vtp-volumetric-transformer-for-multi-view","title":"VTP: Volumetric Transformer for Multi-view Multi-person 3D Pose Estimation","date":"2022-05-25","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/mug-multi-human-graph-network-for-3d-mesh","title":"MUG: Multi-human Graph Network for 3D Mesh Reconstruction from 2D Pose","date":"2022-05-25","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/permutation-invariant-relational-network-for","title":"Permutation-Invariant Relational Network for Multi-person 3D Pose Estimation","date":"2022-04-11","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/distribution-aware-single-stage-models-for","title":"Distribution-Aware Single-Stage Models for Multi-Person 3D Pose Estimation","date":"2022-03-15","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":3,"samples_ran":1,"samples_unverified":2,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/6d-rotation-representation-for-unconstrained","title":"6D Rotation Representation For Unconstrained Head Pose Estimation","date":"2022-02-25","rows_on_this_dataset":1,"code_links":2,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":8,"samples_ran":5,"samples_unverified":3,"pointer_only_for_licence":3,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/direct-multi-view-multi-person-3d-pose","title":"Direct Multi-view Multi-person 3D Pose Estimation","date":"2021-11-07","rows_on_this_dataset":1,"code_links":2,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":1,"samples_ran":1,"samples_unverified":0,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/graph-based-3d-multi-person-pose-estimation","title":"Graph-Based 3D Multi-Person Pose Estimation Using Multi-View Images","date":"2021-09-13","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":22,"samples_ran":19,"samples_unverified":3,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/voxeltrack-multi-person-3d-human-pose","title":"VoxelTrack: Multi-Person 3D Human Pose Estimation and Tracking in the Wild","date":"2021-08-05","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/tessetrack-end-to-end-learnable-multi-person","title":"TesseTrack: End-to-End Learnable Multi-Person Articulated 3D Pose Tracking","date":"2021-06-16","rows_on_this_dataset":4,"code_links":0,"syntology":null},{"paper":"/paper/body-meshes-as-points","title":"Body Meshes as Points","date":"2021-05-06","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":7,"samples_ran":1,"samples_unverified":6,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/multi-view-multi-person-3d-pose-estimation","title":"Multi-View Multi-Person 3D Pose Estimation with Plane Sweep Stereo","date":"2021-04-06","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":9,"samples_ran":6,"samples_unverified":3,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/adafuse-adaptive-multiview-fusion-for","title":"AdaFuse: Adaptive Multiview Fusion for Accurate Human Pose Estimation in the Wild","date":"2020-10-26","rows_on_this_dataset":1,"code_links":2,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":3,"samples_ran":3,"samples_unverified":0,"pointer_only_for_licence":3,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/centerhmr-a-bottom-up-single-shot-method-for","title":"Monocular, One-stage, Regression of Multiple 3D People","date":"2020-08-27","rows_on_this_dataset":1,"code_links":2,"syntology":null},{"paper":"/paper/smap-single-shot-multi-person-absolute-3d","title":"SMAP: Single-Shot Multi-Person Absolute 3D Pose Estimation","date":"2020-08-26","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/hmor-hierarchical-multi-person-ordinal","title":"HMOR: Hierarchical Multi-Person Ordinal Relations for Monocular Multi-Person 3D Pose Estimation","date":"2020-08-01","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/multi-person-3d-pose-estimation-in-crowded","title":"Multi-person 3D Pose Estimation in Crowded Scenes Based on Multi-View Geometry","date":"2020-07-21","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":19,"samples_ran":0,"samples_unverified":19,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/whenet-real-time-fine-grained-estimation-for","title":"WHENet: Real-time Fine-Grained Estimation for Wide Range Head Pose","date":"2020-05-20","rows_on_this_dataset":1,"code_links":4,"syntology":null},{"paper":"/paper/end-to-end-estimation-of-multi-person-3d","title":"VoxelPose: Towards Multi-Camera 3D Human Pose Estimation in Wild Environment","date":"2020-04-13","rows_on_this_dataset":1,"code_links":2,"syntology":null},{"paper":"/paper/light3dpose-real-time-multi-person-3d","title":"Light3DPose: Real-time Multi-Person 3D PoseEstimation from Multiple Views","date":"2020-04-06","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/compressed-volumetric-heatmaps-for-multi","title":"Compressed Volumetric Heatmaps for Multi-Person 3D Pose Estimation","date":"2020-04-01","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/190505754","title":"Learnable Triangulation of Human Pose","date":"2019-05-14","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":17,"samples_ran":1,"samples_unverified":16,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}}],"syntology_totals":{"read_at":"2026-09-24T18:15:14+00:00","papers_with_samples":10,"samples_harvested":93,"samples_ran":37,"samples_unverified":56,"pointer_only_for_licence":10,"papers_with_no_sample_that_ran":2,"note":"the per-paper counts above, summed; not a rate"},"papers_note":"The archive never published its papers-using-dataset list; these are papers with a leaderboard row on this dataset's benchmarks."}