Upload README.md with huggingface_hub

README.md

@@ -38,23 +38,13 @@ Metrics are computed on the held-out test shots in the normalized domain. If a b
 
 ## Model Architectures
 
-- **Attention UNet** (`atten_unet`) — U-Net with additive attention gates on skip connections (Oktay et al., 2018, MIDL). Base channels: 32, depth: 4.
-- **Attention UNet-Plus** (`atten_unet_plus`) — Wider Attention UNet variant. Base channels: 64, depth: 4.
-- **DnCNN** (`dncnn`) — Flat 17-layer Conv-BN-ReLU stack with residual learning (Zhang et al., 2017, IEEE TIP). Base channels: 64.
-- **ResUNet** (`res_unet`) — U-Net with residual blocks replacing plain double-conv layers (He et al., 2016; Zhang et al., 2018). Base channels: 32, depth: 4.
-- **ResUNet-Plus** (`res_unet_plus`) — Wider ResUNet variant. Base channels: 64, depth: 4.
-- **UNet** (`unet`) — Classic encoder-decoder with skip connections (Ronneberger et al., 2015). Base channels: 32, depth: 4.
-- **UNet-Plus** (`unet_plus`) — Wider UNet variant. Base channels: 64, depth: 4.
+- **DNNDAT** (`dnndat`) — DNNDAT-style convolutional encoder-decoder for marine multiple suppression (Wang et al., 2022). U-Net-like encoder-decoder with 28 convolutional layers and dropout.
+- **SAGAN** (`sagan`) — Self-attention GAN generator for seismic surface-related multiple suppression (Tao et al., 2022). U-Net generator with a bottleneck self-attention block.
 
 ## Uploaded Checkpoints
 
-  - Attention UNet: 3 checkpoints
-  - Attention UNet-Plus: 3 checkpoints
-  - DnCNN: 3 checkpoints
-  - ResUNet: 3 checkpoints
-  - ResUNet-Plus: 3 checkpoints
-  - UNet: 3 checkpoints
-  - UNet-Plus: 3 checkpoints
+  - DNNDAT: 4 checkpoints
+  - SAGAN: 3 checkpoints
 
 ### Preprocessing