Open Seyoung9304 opened 2 years ago
All stages are differentiable End-to-end trainable architecture
Feature encoder Extracts per-pixel features from input images (I0, I1) Performed once
Context encoder Extracts per-pixel features from input images (I1) Performed once
Correlation layer
Update operator Update operator estimates {𝑓_0,𝑓_1,𝑓_2,𝑓_3, …,𝑓_𝑁 } from an initial starting point 𝑓_0=0
Abstract