核心概念
Proposing a novel edge computing system for real-time video analysis with adaptive spatial-temporal semantic filtering to maximize processing rate and accuracy.
统计
"DCRL improves the processing rate by up to 66.3% compared to F-B."
"TAODM first observes the state st ∈ St from the environment and selects the best action a∗ t ∈ At for maximizing the action-value function."
"For each context e ∈ E and configuration g ∈ G, we define MAB reward Re,g as Re,g = R · 1(et = e ∧ gt = g)."
引用
"A novel edge computing enabled real-time video analysis system based on spatial-temporal semantic filtering is proposed."
"Simulations are conducted to examine the performance of the proposed DCRL framework on the multi-camera pedestrian video dataset."