Poseidon (✈️ CoRL)

Our biggest speech data release yet:

Today, we’re introducing our Indic Speech Release on @huggingface: 3,845+ hours of transcribed speech across Hindi, Telugu, Tamil, and Bengali.

• 353K+ recordings
• Longer, real-world speech samples
• Speaker + demographic metadata
• Per-example validation signals

Built to help teams train, evaluate, and find the failures that short-clip benchmarks tend to hide.

Come say hello to the team at @corl_conf.

@sebliu will present our paper "When Crowdsourced Egocentric Video Goes Wrong: Failure Modes, Metadata, and Quality Control" at the "Oops, I Erred" workshop.

🗓️ November 12
📍JW Marriott Austin

Inside Poseidon's annotation workflow.

We're building an in-house platform for processing and annotating real-world video, with:

• Frame-level controls and timeline segmentation
• Task-specific annotation fields
• Review and adjudication workflows
• AI-assisted annotation

A quick look at what we're building.

Data in the Wild, ep. 005.

Real-world tasks rarely happen from a fixed viewpoint. This one captures manipulation in a constrained space, with frequent camera movement, occlusion, and re-entry as the task progresses.

Task context
• Clean one area of a car interior + return supplies
• 15 minutes
• A parked car in residential United Kingdom

Capture
• Handheld monocular Infinix X665B
• 1280×720, H.264

Embodied behavior
• Cloth, bucket, dashboard, controls, storage, console + seats
• Actions: rinse_cloth, wring_cloth, wipe_dashboard, wipe_control, open_storage, wipe_console, wipe_seat_gap, wipe_sill
• Challenges: constrained reach, shifting viewpoints + repeated exits

Video sourced from Numo, which collects real-world data that Poseidon then refines into training-ready datasets.

See you at @corl_conf 2026.

Our paper, "When Crowdsourced Egocentric Video Goes Wrong: Failure Modes, Metadata, and Quality Control," was accepted to the "Oops, I Erred" workshop at CoRL, a leading robotics research conference.

If you'll be in Austin, say hi: @sebliu + team will present our research on a critical challenge for Physical AI: how to ensure crowdsourced, real-world video data is reliable and useful for training robots.

corl.org