Question about "structured latent codes"

menyifang / MIMO

Official implementation of "MIMO: Controllable Character Video Synthesis with Spatial Decomposed Modeling"

Apache License 2.0

708 stars 33 forks source link

Open freemantom opened 14 hours ago

freemantom commented 14 hours ago

Are the structured latent codes extracted from the input image sequences, or are they learnable embeddings?