Robots are going to be a big part of our lives in the future. There are plenty of them that can handle household items but teaching them to adapt to new homes can be a challenge. Helix 2.5 aims to solve that. It is a new AI model for robots for household task automation. It enabled robots to perform 3 long-horizon behaviors across 30 unseen homes, without data collection and fine-turning.
More like this ➡️ here
Helix 2.5 30-Home Generalization
As the good folks at Figure explain:
- Behavior specification got 2x cheaper while its scope expanded 30x. Helix 2.5 used half as much task-specific data as a representative Helix 02 behavior, then generalized that behavior across 30 unseen homes.
The robots were tasked with tidying a room, making a bed, and folding towel. Helix 2.5 achieves these in homes never encountered before with a zero-shot approach.
[HT]
*Our articles may contain aff links. As an Amazon Associate we earn from qualifying purchases. Please read our disclaimer on how we fund this site.
