TartanIMU: a light IMU foundation model for robotics, 36% ATE gain
rsasaki0109 · x · 2026-09-08
A CVPR 2025 paper releases TartanIMU, a light "foundation model" for inertial positioning in robotics, with open-source code.
- Predicts 3D body-frame velocity from 6-axis accelerometer + gyroscope data for inertial positioning
- Shared representation across 4 platforms (car, quadruped, drone, human) pretrained on 100+ hours of data
- Reports 36% ATE improvement and 200 FPS online adaptation
- ResNet-LSTM multi-head architecture with pretrained inference, configurable training/eval, and an IROS 2026 challenge starter kit (public release is the LSTM model; Transformer kept for compatibility)
More from Embodied
- Meta faces new lawsuit alleging 'perv glasses' footage was used to train AI without disclosure — jonerp · 2026-09-08
- $995 robot controlled by GPT-6 Astra zero-shots opening a cabinet door and self-corrects — DeryaTR_ · 2026-09-08
- GPT-6 Demo Shows End-to-End Real2Sim From Robot Footage in One Prompt — CyberRobooo · 2026-09-08
- One-shotting a Raspberry Pi UPS HAT CAD model with Claude in a few minutes — burhop · 2026-09-08
- Tesla Cybercab caught driving calmly in heavy rain, wipers stay relaxed — yunta_tsai · 2026-09-08
- AI-powered cooking robots deployed in 4,000 stores across 300+ Chinese cities — Polymarket · 2026-09-08