SOTAVerified

Pose Estimation

Pose Estimation is a computer vision task where the goal is to detect the position and orientation of a person or an object. Usually, this is done by predicting the location of specific keypoints like hands, head, elbows, etc. in case of Human Pose Estimation.

A common benchmark for this task is MPII Human Pose

( Image credit: Real-time 2D Multi-Person Pose Estimation on CPU: Lightweight OpenPose )

Papers

Showing 651700 of 4228 papers

TitleStatusHype
Learning to Predict 3D Lane Shape and Camera Pose from a Single Image via Geometry ConstraintsCode1
AdaptivePose: Human Parts as Adaptive PointsCode1
Rethinking the Data Annotation Process for Multi-view 3D Pose Estimation with Active Learning and Self-TrainingCode1
SmoothNet: A Plug-and-Play Network for Refining Human Poses in VideosCode1
AdaptPose: Cross-Dataset Adaptation for 3D Human Pose Estimation by Learnable Motion GenerationCode1
Space Non-cooperative Object Active Tracking with Deep Reinforcement LearningCode1
PeopleSansPeople: A Synthetic Data Generator for Human-Centric Computer VisionCode1
CrossLoc: Scalable Aerial Localization Assisted by Multimodal Synthetic DataCode1
DProST: Dynamic Projective Spatial Transformer Network for 6D Pose EstimationCode1
ElePose: Unsupervised 3D Human Pose Estimation by Predicting Camera Elevation and Learning Normalizing Flows on 2D PosesCode1
Local and Global Point Cloud Reconstruction for 3D Hand Pose EstimationCode1
PE-former: Pose Estimation TransformerCode1
ScaleNet: A Shallow Architecture for Scale EstimationCode1
ALIKE: Accurate and Lightweight Keypoint Detection and Descriptor ExtractionCode1
Robust Pose Estimation in Crowded Scenes with Direct Pose-Level InferenceCode1
HRFormer: High-Resolution Vision Transformer for Dense PredictCode1
Camera Distortion-aware 3D Human Pose Estimation in Video with Optimization-based Meta-LearningCode1
MHFormer: Multi-Hypothesis Transformer for 3D Human Pose EstimationCode1
VideoPose: Estimating 6D object pose from videosCode1
Rethinking Keypoint Representations: Modeling Keypoints and Poses as Objects for Multi-Person Human Pose EstimationCode1
IKEA Object State Dataset: A 6DoF object pose estimation dataset and benchmark for multi-state assembly objectsCode1
Co-segmentation Inspired Attention Module for Video-based Computer Vision TasksCode1
Sparse Steerable Convolutions: An Efficient Learning of SE(3)-Equivariant Features for Estimation and Tracking of Object Poses in 3D SpaceCode1
TANDEM: Tracking and Dense Mapping in Real-time using Deep Multi-view StereoCode1
Direct Multi-view Multi-person 3D Pose EstimationCode1
ROFT: Real-Time Optical Flow-Aided 6D Object Pose and Velocity TrackingCode1
Efficient 3D Deep LiDAR OdometryCode1
HHP-Net: A light Heteroscedastic neural network for Head Pose estimation with uncertaintyCode1
3DP3: 3D Scene Perception via Probabilistic ProgrammingCode1
Neural View Synthesis and Matching for Semi-Supervised Few-Shot Learning of 3D PoseCode1
SRT3D: A Sparse Region-Based 3D Object Tracking Approach for the Real WorldCode1
Occlusion-Robust Object Pose Estimation with Holistic RepresentationCode1
MOS: A Low Latency and Lightweight Framework for Face Detection, Landmark Localization, and Head Pose EstimationCode1
HRFormer: High-Resolution Transformer for Dense PredictionCode1
TransFusion: Cross-view Fusion with Transformer for 3D Human Pose EstimationCode1
HUMAN4D: A Human-Centric Multimodal Dataset for Motions and Immersive MediaCode1
The Center of Attention: Center-Keypoint Grouping via Attention for Multi-Person Pose EstimationCode1
BEV-Net: Assessing Social Distancing Compliance by Joint People Localization and Geometric ReasoningCode1
Towards Accurate Cross-Domain In-Bed Human Pose EstimationCode1
SPEED+: Next-Generation Dataset for Spacecraft Pose Estimation across Domain GapCode1
Keypoint CommunitiesCode1
Hierarchical Kinematic Probability Distributions for 3D Human Shape and Pose Estimation from Images in the WildCode1
PDC-Net+: Enhanced Probabilistic Dense Correspondence NetworkCode1
PASS: An ImageNet replacement for self-supervised pretraining without humansCode1
Multi-View Video-Based 3D Hand Pose EstimationCode1
The Hilti SLAM Challenge DatasetCode1
AirDOS: Dynamic SLAM benefits from Articulated ObjectsCode1
BabelCalib: A Universal Approach to Calibrating Central CamerasCode1
Unsupervised 3D Pose Estimation for Hierarchical Dance Video RecognitionCode1
HPTQ: Hardware-Friendly Post Training QuantizationCode1
Show:102550
← PrevPage 14 of 85Next →

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1yoloposeAP5090.3Unverified
2ViTPose (ViTAE-G, ensemble)AP81.1Unverified
3ViTPose (ViTAE-G)AP80.9Unverified
4PoseBH-HAP79.5Unverified
5UDP-Pose-PSA(384x288)AP79.5Unverified
64xRSN-50 (ensemble)AP79.2Unverified
7UDP-Pose-PSA(256x192)AP78.9Unverified
8CCM+AP78.9Unverified
94xRSN-50AP78.6Unverified
10PCT (256x256)AP78.3Unverified
#ModelMetricClaimedVerifiedStatus
1PCT (swin-l, test set)PCKh-0.594.3Unverified
2Soft-gated Skip ConnectionsPCKh-0.594.1Unverified
3Cascade Feature AggregationPCKh-0.593.9Unverified
4PCT (swin-b, test set)PCKh-0.593.8Unverified
5TransPosePCKh-0.593.5Unverified
6UniHCP (FT)PCKh-0.593.2Unverified
74xRSN-50PCKh-0.593Unverified
8UniPosePCKh-0.592.7Unverified
9MSPNPCKh-0.592.6Unverified
10Spatial ContextPCKh-0.592.5Unverified
#ModelMetricClaimedVerifiedStatus
1ViTPose (ViTAE-G, GT bounding boxes)Test AP93.3Unverified
2UniHCP (direct eval)Test AP87.4Unverified
3PoseBH-HTest AP87Unverified
4RTMPose(RTMPose-l, GT bounding boxes)Test AP80.3Unverified
5TransPose-HValidation AP62.3Unverified
6BBox-Mask-Pose 2xTest AP48.3Unverified
7BUCTD (CID-W32)Test AP47.2Unverified
8HQNet (ViT-L)Test AP45.6Unverified
9MaskPose-bTest AP45Unverified
10CID (HRNet-W48)Test AP45Unverified
#ModelMetricClaimedVerifiedStatus
1OmniPosePCK99.5Unverified
2Soft-gated Skip ConnectionsPCK94.8Unverified
3Residual Hourglass + ASR + AHOPCK94.5Unverified
4UniPosePCK94.5Unverified
5Chou et al. arXiv'17PCK94Unverified
6Pyramid Residual Modules (PRMs)PCK93.9Unverified
7Stacked hourglass + Inception-resnetPCK93.9Unverified
8Multi-Context AttentionPCK92.6Unverified
9FPDPCK90.8Unverified
10Part heatmap regression (ResNet-152)PCK90.7Unverified
#ModelMetricClaimedVerifiedStatus
1BUCTD-W48 (w/cond. input from PETR, and generative sampling)AP78.5Unverified
2ViTPose-GAP78.3Unverified
3BUCTD-W48 (w/cond. input from PETR)AP76.7Unverified
4SwinV2-L 1K-MIMAP75.5Unverified
5SwinV2-B 1K-MIMAP74.9Unverified
6BUCTD-W48AP72.9Unverified
7OpenPifPafAP70.5Unverified
8MIPNet (HRNet-W48)AP70Unverified
9KAPAO-LAP68.9Unverified
10KAPAO-MAP67.1Unverified
#ModelMetricClaimedVerifiedStatus
1CCNet (ViTPose-B_GT-bbox_256x192)AP78.1Unverified
2MogaNet-B (384x288)AP77.3Unverified
3ViTPose-B (Single-task_GT-bbox_256x192)AP77.3Unverified
4MogaNet-S (384x288)AP76.4Unverified
5Bias (HRNet_256x192)AP75.8Unverified
6ViTPose-B (Single-task_Det-bbox_256x192)AP75.8Unverified
7HRNet (256x192)AP75.3Unverified
8MogaNet-S (256x192)AP74.9Unverified
9MogaNet-T (256x192)AP73.2Unverified
10RLE (256x192)AP71.3Unverified
#ModelMetricClaimedVerifiedStatus
1Hulk(Finetune, ViT-L)AP37.1Unverified
2Hulk(Finetune, ViT-B)AP35.6Unverified
3HRFormer (HRFomer-B)AP34.4Unverified
4UniHCP (finetune)AP33.6Unverified
5HRNet (HRNet-w48 )AP33.5Unverified
6HRNet (HRNet-w32)AP32.3Unverified
7HRFormer (HRFomer-S)AP31.6Unverified
8SimpleBaseline (ResNet-152)AP29.9Unverified
9SimpleBaseline (ResNet-101)AP29.4Unverified
10SimpleBaseline (ResNet-50)AP28Unverified
#ModelMetricClaimedVerifiedStatus
1BUCTD (PETR, with generative sampling)APL83.7Unverified
2OmniPose (WASPv2)AP79.5Unverified
3MetaPrompt-SDAP79Unverified
4Hulk(Finetune, ViT-L)AP78.7Unverified
5BUCTD (PETR, with generative sampling)AP77.8Unverified
6Hulk(Finetune, ViT-B)AP77.5Unverified
7I²R-Net (1st stage:HRFormer-B)AP77.3Unverified
8PATH (Partial FT)AP77.1Unverified
9SOLIDER (swin-B)AP76.6Unverified
10PEFORMER-Xcit-dino-p8AP72.6Unverified
#ModelMetricClaimedVerifiedStatus
1GIM-DKMDUC1-Acc@0.25m,10°57.1Unverified
2GIM-LoFTRDUC1-Acc@0.25m,10°54.5Unverified
3GIM-SuperGlueDUC1-Acc@0.25m,10°53.5Unverified
4DKMDUC1-Acc@0.25m,10°51.5Unverified
5SuperGlueDUC1-Acc@0.25m,10°49Unverified
6LoFTRDUC1-Acc@0.25m,10°47.5Unverified
#ModelMetricClaimedVerifiedStatus
1AdaPoseMean mAP93.38Unverified
2DECA-D3Mean mAP88.75Unverified
3V2V-PoseNetMean mAP88.74Unverified
4A2JMean mAP88Unverified
5RENMean mAP84.9Unverified
6Multi-task learning + viewpoint invarianceMean mAP77.4Unverified
#ModelMetricClaimedVerifiedStatus
1SimpleBaseline + HANetMean PCK@0.299.6Unverified
2DeciWatchMean PCK@0.299Unverified
3LSTM PMMean PCK@0.293.6Unverified
4CPMMean PCK@0.291.9Unverified
5UniTrack_i18Mean PCK@0.280.5Unverified
#ModelMetricClaimedVerifiedStatus
14xRSN-50PCKh@0.593Unverified
2RefinePCKh@0.592.1Unverified
3EfficientPose IVPCKh@0.591.2Unverified
4OpenPosePCKh@0.588.8Unverified
5Adversarial LearningPCKh@0.588.6Unverified
#ModelMetricClaimedVerifiedStatus
1OmniPoseMean PCK@0.299.4Unverified
2UniPose-LSTMMean PCK@0.299.3Unverified
3LSTM PMMean PCK@0.297.7Unverified
4Thin-SlicingMean PCK@0.296.5Unverified
5Iqbal et al.Mean PCK@0.281.1Unverified
#ModelMetricClaimedVerifiedStatus
1DP-RCNN-DeepLab (ResNet-101)AP68Unverified