iug8oyo8mit
IMS2026 HW0 — media
Media artifacts including videos, checkpoints, and evaluation results for vision-language-action robotics experiments using TIC-VLA on DynaNav navigation tasks and SmolVLA on LIBERO manipulation tasks in Isaac Sim.
Downloads46
Episodes408
Why This Matters for Physical AI
Demonstrates vision-language-action model performance on navigation and manipulation tasks across different robot platforms and environments, providing benchmarks for embodied AI policy evaluation.
Technical Profile
- Modalities
- rgblanguage
- Robot Embodiments
- Nova CarterSpot
- Environment
- simulationhospitalwarehouseofficeoutdoor
- Task Types
- navigationmanipulationpick_and_place
- Episodes
- 408
- Data Format
- H.264
- Annotation Types
- language_instructions
- License
- mit
Access
Need custom rgb data?
Claru builds purpose-built datasets for simulation applications with dense human annotations and quality assurance.
Request a Sample Pack