Skinned Multi Person Linear Model
PulseAugur coverage of Skinned Multi Person Linear Model — every cluster mentioning Skinned Multi Person Linear Model across labs, papers, and developer communities, ranked by signal.
3 day(s) with sentiment data
-
AI system wins Parkinsonian gait challenge with novel aggregation techniques
Researchers developed a system that won the MoCha 2026 Benchmark and Challenge on Parkinsonian Gait by predicting gait severity from motion data. The system achieved a macro-F1 score of 0.6945 on the hidden test set, ou…
-
New benchmarks and datasets advance human motion tracking and generation
Researchers have introduced new benchmarks and datasets for evaluating human motion tracking and generation. HiPHI offers over 600 hours of high-fidelity motion data, guided by linguistic principles, to improve humanoid…
-
New volumetric body model improves fetal MRI analysis with KTPolyRigid transform
Researchers have developed a new differentiable volumetric body model for medical imaging analysis, utilizing a novel Kinematic Tree-based Log-Euclidean PolyRigid (KTPolyRigid) transform. This approach addresses limitat…
-
New T3HG-Editor enables text-driven 3D human garment editing
Researchers have introduced T3HG-Editor, a novel system for text-driven 3D human garment editing. This approach addresses limitations in existing 3D Gaussian Editing methods, which often produce low-fidelity and inconsi…
-
Superman framework unifies human motion perception and generation
Researchers have introduced Superman, a novel framework designed to unify human motion perception and generation tasks. This system bridges the gap between understanding motion from video and generating temporal skeleto…
-
Patient-specific articulated digital twins created from single CT scans
Researchers have developed a method to create patient-specific articulated digital twins from a single full-body CT scan. This technique fits a parametric human body model to establish a kinematic scaffold, then binds s…
-
New framework estimates 3D human motion and kinetics from monocular video
Researchers have developed MonoMSK, a novel framework for estimating 3D human motion and kinetics from monocular video. This hybrid approach combines data-driven learning with physics-based simulation, utilizing an anat…
-
PoseShield tackles human self-collision in pose estimation
Researchers have developed PoseShield, a novel method to address self-collision issues in human pose estimation and motion generation. This technique defines a neural collision constraint directly within the SMPL pose s…
-
EMOSH framework disentangles motion and shape for expressive human animation
Researchers have developed EMOSH, a new framework for generating high-fidelity human animations that addresses the issue of motion-shape entanglement. EMOSH introduces an Expressive Human Model (EHM) that explicitly dis…
-
TopoCap framework animates any 3D character from video
Researchers have developed TopoCap, a novel framework for generating animations from monocular video that can adapt to any skeletal structure. This system learns a universal motion manifold, disentangling motion dynamic…
-
EgoPriMo framework generates humanoid robot motion from human demos
Researchers have developed EgoPriMo, a new framework for generating full-body motion for humanoid robots using egocentric human demonstrations. This system takes egocentric visual observations and text prompts to recons…
-
New methods generate realistic 3D avatars from single photos
Researchers have developed two new methods for generating realistic 3D human avatars from single images. Splatshot combines 3D Gaussian Splatting with diffusion models to ensure both geometric consistency and photoreali…
-
MotionPRO dataset enhances human motion capture with pressure data
Researchers have developed MotionPRO, a new dataset and methodology for human motion capture that incorporates pressure sensor data alongside traditional visual and optical sensors. This approach aims to improve the phy…
-
New method quantifies and reduces noise in video-based joint torque estimation
Researchers have developed a new method to analyze and mitigate noise amplification in video-based joint torque estimation. Their findings indicate that pose estimation noise can be amplified by approximately 1,000 time…
-
AI detects suggestive motion in 3D environments using skeleton data
Researchers have developed a new AI pipeline to detect suggestive and explicit movements in 3D virtual environments using skeleton data. The system analyzes motion fragments based on Laban Movement Analysis descriptors,…
-
mmWave Radar Scans Offer Non-Intrusive Body Composition Assessment
Researchers have developed new methods for assessing body composition using millimeter-wave (mmWave) radar scans, which can penetrate clothing and preserve privacy. One approach uses multi-task learning to predict visce…
-
Researchers develop new generative model for realistic human geometry and clothing details
Researchers have developed a novel generative model for human geometry that significantly improves the quality and efficiency of creating realistic 3D avatars. This new approach encodes geometry distributions as 2D feat…
-
Action Motifs paper introduces self-supervised hierarchical human movement representation
Researchers have developed a novel hierarchical representation for human body movements called Action Motifs. This system uses Action Atoms to capture atomic joint movements and Action Motifs to encode temporal composit…
-
ETCH-X and OmniFit advance 3D clothed human body fitting with new methods
Researchers have developed ETCH-X, an advanced method for fitting parametric body models to 3D scans of clothed humans. This new approach improves upon its predecessor by incorporating a "tightness-aware" fitting paradi…