Table 2: Detailed Information of SDF Autoencoder

Encoder
Layer 1	conv(1, 32) + conv(32, 32) + pooling(64 × 64 × 64 × 1 → 32 × 32 × 32 × 32)
Layer 2	conv(32, 64) + conv(64, 64) + pooling(32 × 32 × 32 × 32 → 16 × 16 × 16 × 64)
Layer 3	conv(64, 128) + conv(128, 128) + pooling(16 × 16 × 16 × 64 → 8 × 8 × 8 × 128)
Layer 4	conv(128, 128) + conv(128, 128) + fc(65536, 256) + fc(256, 256)(8 × 8 × 8 × 128 → 256)
Decoder
Layer 1	fc(256, 256) + fc(256, 65536) + conv(128, 128) + conv(128, 128)(256 → 8 × 8 × 8 × 128)
Layer 2	deconv(128, 128) + conv(128, 128) + conv(128, 64)(8 × 8 × 8 × 128 → 16 × 16 × 16 × 64)
Layer 3	deconv(64, 64) + conv(64, 64) + conv(64, 32)(16 × 16 × 16 × 64 → 32 × 32 × 32 × 32)
Layer 4	deconv(32, 32) + conv(32, 32) + conv(32, 1)(32 × 32 × 32 × 32 → 64 × 64 × 64 × 1)