SOURCE-LINKED INTELLIGENCE

Input Resolution Matters: Real-Time Object Detection Latency

arXiv · AI, language, vision and robotics · article · Sep 11, 2026 · UTC

We model total latency as the convolution of preprocessing, inference, and postprocessing distributions under a simplifying independence approximation, with selected stage parameters expressed as functions of source-image resolution. Under this assumption, the probability density of the total latency is the convolution of the stage-wise densities, and its cumulative distribution function (CDF) provides the distribution of end-to-end detection time. Each stage is modeled by a parametric distribution (e.g., Exponential, Erlang, Normal, Gamma), with parameters expressed as functions of the source

Read original source ↗ Open in workspace

recordType

paper

region

Global