Train the ResNet Feat Extractor Backbone From Scratch?

google / dynibar

Implementation of DynIBaR Neural Dynamic Image-Based Rendering (CVPR 2023)

https://dynibar.github.io/

Apache License 2.0

852 stars 40 forks source link

Closed GloryyrolG closed 7 months ago

GloryyrolG commented 7 months ago

Hi @zhengqili @googlebot et al.,

Thx for ur impressive contribution. May I ask if

means we do not use a pre-trained ResNet? Thx & best,

GloryyrolG commented 7 months ago

It is no different from training with/ a pre-trained ResNet checkpoint.