S2A2 adds microphone-array spatial audio and spectrograms to imitation-learning policies, substantially improving success on manipulation tasks where vision alone cannot identify the target or destination.
Title resolution pending
1 Pith paper cite this work, alongside 71 external citations. Polarity classification is still indexing.
1
Pith paper citing it
71
external citations · OpenAlex
fields
cs.RO 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
S2A2: Audio-Visual Imitation Learning for Manipulation Tasks Using Acoustic Spatial Information
S2A2 adds microphone-array spatial audio and spectrograms to imitation-learning policies, substantially improving success on manipulation tasks where vision alone cannot identify the target or destination.