LIST3R reconnects fragmented video subsequences using persistent instance anchors with semantic and geometric evidence to produce consistent global 3D reconstructions.
Entitysam: Segment everything in video
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
LIST3R: Long-sequence Instance-aware 3D Reconstruction
LIST3R reconnects fragmented video subsequences using persistent instance anchors with semantic and geometric evidence to produce consistent global 3D reconstructions.