Tactile Module
Records high-precision, distributed tactile interaction data.
Data Collection Solutions for Embodied AI
PaXini turns human experience into high-quality, reusable data for Physical AI. Our integrated hardware, software and processing pipeline captures multimodal data—from precise hand and upper-body movements in tabletop tasks to full-body activity in open-world environments.
PXCap Collection Device
PaXini's PXCap product series provides a complete, open-architecture hardware system for seamless data collection across diverse environments.
OmniSharing DB
PaXini provides multimodal data spanning multidimensional tactile, multi-view vision, speech, text, proprioception and spatial trajectories, enabling embodied AI to generalize rapidly across scenarios.
The Way we Collect
PaXini offers In Domian and In the Wild data collection solutions to meet embodied AI's data needs across diverse scenarios.
Dataset Features
Comprehensive data diversity for embodied AI development
Joint Module
Records joint angles from the collection glove and tracks global trajectory data.
Visual Module
Provides synchronized multi-view, high-frame-rate video streams that comprehensively capture the task scenario.
Spatial Trajectory
Process multi-view synchronized video streams and annotate corresponding spatial trajectories using estimation algorithms.
Voice Module
Voice data collected before task execution, describing the task.
Text Module
Structured textual information obtained by converting the collected Voice data.
PXCap Pro
A lightweight data glove for scalable collection, natural operation and comfortable wear. Multidimensional tactile sensing, magnetic joint encoders and wrist camera capture detailed hand-object interactions.
17 DoFLightweight and Easy to UseWrist ViewMultidimensional Tactile Sensing
PXCap III&PXDex III
The capture glove and three-finger end effector share a 1:1 physical design, with closely matched sensor layouts and kinematics for direct human-to-robot data mapping.
Isomorphic Capture and Execution1:1 Motion MappingDirect Data TransferHigh-Fidelity Reproduction
PXCap Ultra
Multidimensional tactile sensors and ultra-compact magnetic encoders provide full-hand coverage for precise capture of joint motion, contact states and force changes.
29 DoF15 Multidimensional Tactile SensorsUltra-Compact Magnetic EncodersFull-Hand Sensing
PXCap Ego
A head-mounted multimodal device combining RGB, IR and high-precision IMU sensing to synchronously capture first-person environmental views, head motion and spatial pose.
280° Total FOV6-DoF Pose TrackingMultimodal SensingLightweight Wearable Design
PXCap Link
A portable edge computing hub for in-the-wild data collection, integrating computing, power supply and data transfer with on-device parsing, preprocessing, storage and upload.
Modular DesignEdge Computing4-Hour Battery LifePortable Deployment
OmniSharing DB
PaXini's OmniSharing DB is a large-scale, high-quality multimodal dataset captured from real-world human activity, spanning multidimensional tactile data, multi-view vision, speech, text, proprioception and spatial trajectories.
15+N
Scenario Types
1,000+
Subtasks
Data Customization
Multiple Data Collection Solutions
Please download the data sample first, then open Data Demo to view the content.
In-Domain Collection
Targeting dexterous manipulation scenarios, human-centric high-precision sampling generates high-fidelity haptic data to close model data loops.
Operation-focusedHuman-CentricPrecision-guaranteed
In the Wild Collection
Open collection builds upon In the Wild data collection, relaxes scenario constraints to enrich the model’s environmental awareness, improves long-sequence task planning performance, and evolves iteratively to expand diversified data collection solutions.
Egocentric DataLow-Cost Large-Scale CollectionReal-Environment Collection
Scene Types
15+N
Task Categories
100+
Task Categories
1000+
All Scenes
Office
Home
Healthcare
Medical
Rehabilitation
Logistics
Education
Entertainment
3C Manufacturing
Restaurant
Agriculture
Business
Retail
Disaster Rescue
Automotive
Industrial
Meta Action
Data Generalization
Cross-embodiment capabilities and comprehensive data characteristics
Data Features
Simulation Env
High-Fidelity physical sensing&Sim Platform
Digital Assets
High Fidelity 3D Gasuusian splatting
Multi-Adaptability
High-Accuracy Retargeting&Response
Different Method
Comparing different data collection approaches for embodied AI
Data Pipeline
How We Design Our Data Pipeline and Process Data
Omnisharing Toolkit
How We Design Our Data Pipeline and Process Data
PX Pose
the pose estimation and processing module
Raw visual data parsing
DF-1 format parses RGBD camera information and aligns multi camera temporal data to provide input for pose estimation.
Pose Estimation of Wristbands and Objects
Binocular infrared depth estimation, wristband automatically detects pose, and objects need to be manually labeled with masks to estimate pose.
PX Post-Process
Parses and converts after all raw data is obtained
Parsing
Processes the raw data output from PX Pose and generates the DF-2 data format.
Retargeting
Retargets DF-2 data to the embodiment and outputs DF-2R.
Conversion
Processes DF-2R data and exports it in the LeRobot data format.
Data Format
Hierarchical dataset organization across all processing stages
DF-1
The overall input: raw data after preprocessing and quality inspection
DF-2
1st output: DF-1 with encoder and tactile data parsed; adds bimanual and object poses; includes both action and observation
The observation layer in DF-2 and DF-2R is the state data in the episode, while action is one frame behind observation. To ensure both arrays have equal length, the last frame in action is repeated.
DF-2R
2nd output: DF-2 retargeted to a dexterous hand model
The observation layer in DF-2 and DF-2R is the state data in the episode, while action is one frame behind observation. To ensure both arrays have equal length, the last frame in action is repeated.
DF-3
3rd output: converts DF-2R to the LeRobot dataset format; can be used for VLA model training
Dataset Structure DF-1
▼
task_list▼
task_XXX▼
episode_0_HHMMSS_CollectionDate_RoomID_CollectorsID.hdf5▼
dataset-
metaTask Description▼
observation-
audioCompressed Audio Stream, Including Text Description▼
state▼
left_hand▼
encoder-
data-
timestamp▼
tactile-
data-
timestamp▼
right_hand▼
encoder-
data-
timestamp▼
tactile-
data-
timestamp▼
image-
rgbd_rgb_extrinsic▼
RGB_CameraXXX-
timestamp-
data-
extrinsics-
intrinsics▼
RGBD_CameraXXX-
timestamp-
extrinsics-
inner_extrinsics▼
color-
data-
intrinsics▼
depth-
data▼
left-
data-
intrinsics▼
right-
data-
intrinsics-
episode_1_HHMMSS_CollectionDate_RoomID_CollectorsID.hdf5-
episode_2_HHMMSS_CollectionDate_RoomID_CollectorsID.hdf5/
......▼
task_XXX/
......Data Closed Loop
Full-Pipeline Real2Sim2Real Data Capabilities and Models for Closing the Data Loop
Real & Simulated Data Verification
Publications
PaperDatasetCode
OmniVTLA: Vision-Tactile-Language-Action Model with Semantic-Aligned Tactile Sensing
Paper
TacCompress: A Benchmark for Multi-Point Tactile Data Compression in Dexterous Manipulation
PaXini · Beijing
+86 755 2357 4593
PaXini · Shenzhen
+86 755 2357 4593
PaXini · Shanghai
+86 21 5456 1536
PaXini · Tianjin
+86 22 8241 1882
PaXini · Wuhan
+86 27 8309 0662
PaXini · SuQian
+86 527 8080 6580
mkt@paxini.com China
sales_global@paxini.com Overseas
hr@paxini.com