SIR uses learned sparse scene graphs from images as an intermediate representation to improve robot policy success rates on RoboCasa and enable analysis of model decisions for dataset biases.
Collective behavior clone with visual attention via neural interaction graph prediction.arXiv preprint arXiv:2503.06869, 2025
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.RO 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
SIR: Structured Image Representations for Explainable Robot Learning
SIR uses learned sparse scene graphs from images as an intermediate representation to improve robot policy success rates on RoboCasa and enable analysis of model decisions for dataset biases.