Question about The Target-Dependent Update (TDU) unit

YufeiWang777 / LRRU

Official implementation of ``LRRU: Long-short Range Recurrent Updating Networks for Depth Completion'', ICCV 2023.

83 stars 4 forks source link

Question about The Target-Dependent Update (TDU) unit #5

Closed DachunKai closed 11 months ago

DachunKai commented 11 months ago

In the paper, Figure 3's TDU uses target dense depth to extract self-guided features during training. However, during inference, the target dense depth is not available, and I am unsure how it works.