Robot Visual-Tactile Sensor Fusion
Build robot visual-tactile fusion with time alignment, frame calibration, contact state, uncertainty, modality ablation, dropout and control tests.
English articles about robot data and hardware at Physical AI Lab.
Build robot visual-tactile fusion with time alignment, frame calibration, contact state, uncertainty, modality ablation, dropout and control tests.
Coordinate two robot arms with absolute and relative pose, closed-chain constraints, internal force, load sharing, timing, collision and recovery tests.
Design robot in-hand manipulation with rolling, sliding, finger gaiting, tactile feedback, pose tracking, force limits, Sim2Real and recovery tests.
Test vacuum gripper failure from leaks, rough or porous surfaces, pressure buildup, acceleration, eccentric load, pad wear and loss detection.
Design robot peg-in-hole force control with bounded approach, contact detection, compliance, search, insertion, jamming, seating and recovery tests.
Control robot camera rolling shutter and motion blur with readout time, exposure, image speed, lighting, triggers, timestamps and motion tests.
Evaluate robot grasp stability with friction cones, force closure, contact geometry, load, slip sensing, object damage and transport tests.
Compare ICP and NDT registration by correspondences, voxel distributions, initial pose, overlap, scale, degeneracy, scores and validation.
Estimate robot 6D object pose with calibrated RGB or depth, PnP, refinement, symmetry, transforms, timing, uncertainty and grasp tests.
Calibrate wheel odometry, detect slip with IMU and scan residuals, adapt covariance, protect frames and validate surfaces, payloads and recovery.