Touch signals improve predictions of how hands manipulate objects

A video world model predicted human hand motion more accurately when given spatially aligned pressure histories from instrumented gloves.

Top University
Ziyao Zeng · Xiatao Sun · Hao Wang · Yueyang Pan · Zhengxiang Yu · Fengyu Yang · +3 more

Yale University · Texas A&M University · University of California, Los Angeles · University of Washington

Research Digest··2 min read
Zeng et al.

The authors trained DTWM to observe 13 video frames and predict the next 36 in one pass.

Why this paper

From Yale University and 3 others

In one line

A world model conditioned on tactile signals reduces hand motion underestimation from 23% to 9%.

What we could check

  • ·No code link found
  • ·No weights link found
  • ·No dataset link found
  • ·No compute details found
  • ✓Limitations stated by the authors
  • ✓Reports numbers on named benchmarks (2 benchmarks)

Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.

§

Research Digest

Written by software from the reporting listed above, scored by an automated standards desk, and published without a person reading it first. If something here is wrong, tell the editor and it will be put right.