A two-stream multimodal model using Motion-Mamba and Impact-Griffin branches with cross-conditioned fusion achieves 96.1% accuracy on a new bathroom fall dataset while cutting latency and energy use on Raspberry Pi 4B compared to baselines.
Title resolution pending
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
-
Edge-Efficient Two-Stream Multimodal Architecture for Non-Intrusive Bathroom Fall Detection
A two-stream multimodal model using Motion-Mamba and Impact-Griffin branches with cross-conditioned fusion achieves 96.1% accuracy on a new bathroom fall dataset while cutting latency and energy use on Raspberry Pi 4B compared to baselines.