A temporal prior from inter-frame motion flow is injected as initialization into an attention pyramid network to guide coarse-to-fine instrument segmentation in MIS videos, exceeding prior results on the EndoVis dataset across three tasks.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CV 2verdicts
UNVERDICTED 2representative citing papers
Fine-tuned MLLMs achieve competitive skeletal landmark localization on synthetic and real X-ray datasets compared to deep learning baselines and demonstrate reasoning for sequential C-arm navigation.
citing papers explorer
-
Incorporating Temporal Prior from Motion Flow for Instrument Segmentation in Minimally Invasive Surgery Video
A temporal prior from inter-frame motion flow is injected as initialization into an attention pyramid network to guide coarse-to-fine instrument segmentation in MIS videos, exceeding prior results on the EndoVis dataset across three tasks.
-
Autonomous Skeletal Landmark Localization towards Agentic C-Arm Control
Fine-tuned MLLMs achieve competitive skeletal landmark localization on synthetic and real X-ray datasets compared to deep learning baselines and demonstrate reasoning for sequential C-arm navigation.