search

lhoyer / HRDA

[ECCV22] Official Implementation of HRDA: Context-Aware High-Resolution Domain-Adaptive Semantic Segmentation
Other
231 stars 32 forks source link

Questions about model forward #20

Closed JoinWorldMC closed 1 year ago

JoinWorldMC commented 1 year ago

In the figure2(b), HR Detail Crops is input to the encoder. After that, the output of the encoder is input to the decoder. My question is: How many forwards are required to infer a single image?