One hierarchical slot-attention model with 10% labels jointly yields holistic, semantic, and panoptic scene decompositions that outperform three separate flat baselines by large ARI margins.
Learning physical graph repre- sentations from visual scenes.Advances in Neural Informa- tion Processing Systems, 33:6027–6039, 2020
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
HSA: Hierarchical Slot Attention for Multi-granularity Scene-Decomposition
One hierarchical slot-attention model with 10% labels jointly yields holistic, semantic, and panoptic scene decompositions that outperform three separate flat baselines by large ARI margins.