feat: enable to use multiple rgb encoders per camera

lerobot/common/policies/diffusion/configuration_diffusion.py

@@ -130,6 +130,7 @@ class DiffusionConfig:
     pretrained_backbone_weights: str | None = None
     use_group_norm: bool = True
     spatial_softmax_num_keypoints: int = 32
+    rgb_encoder_per_camera: bool = False
     # Unet.
     down_dims: tuple[int, ...] = (512, 1024, 2048)
     kernel_size: int = 5

lerobot/common/policies/diffusion/modeling_diffusion.py

@@ -182,8 +182,14 @@ def __init__(self, config: DiffusionConfig):
         self._use_env_state = False
         if num_images > 0:
             self._use_images = True
-            self.rgb_encoder = DiffusionRgbEncoder(config)
-            global_cond_dim += self.rgb_encoder.feature_dim * num_images
+            if self.config.rgb_encoder_per_camera:
+                encoders = [DiffusionRgbEncoder(config) for _ in range(num_images)]
+                self.rgb_encoder = nn.ModuleList(encoders)
+                global_cond_dim += encoders[0].feature_dim * num_images
+            else:
+                self.rgb_encoder = DiffusionRgbEncoder(config)
+                global_cond_dim += self.rgb_encoder.feature_dim * num_images
+            print(f"gloabl cond dim: {global_cond_dim}")
         if "observation.environment_state" in config.input_shapes:
             self._use_env_state = True
             global_cond_dim += config.input_shapes["observation.environment_state"][0]
@@ -241,14 +247,30 @@ def _prepare_global_conditioning(self, batch: dict[str, Tensor]) -> Tensor:
         global_cond_feats = [batch["observation.state"]]
         # Extract image feature (first combine batch, sequence, and camera index dims).
         if self._use_images:
-            img_features = self.rgb_encoder(
-                einops.rearrange(batch["observation.images"], "b s n ... -> (b s n) ...")
-            )
-            # Separate batch dim and sequence dim back out. The camera index dim gets absorbed into the
-            # feature dim (effectively concatenating the camera features).
-            img_features = einops.rearrange(
-                img_features, "(b s n) ... -> b s (n ...)", b=batch_size, s=n_obs_steps
-            )
+            if self.config.rgb_encoder_per_camera:
+                images_per_camera = einops.rearrange(batch["observation.images"], "b s n ... -> n (b s) ...")
+                img_features_list = torch.cat(
+                    [
+                        encoder(images)
+                        for encoder, images in zip(self.rgb_encoder, images_per_camera, strict=False)
+                    ]
+                )
+                img_features = einops.rearrange(
+                    img_features_list, "(n b s) ... -> b s (n ...)", b=batch_size, s=n_obs_steps
+                )
+                print("using rgb encoder per camera")
+                print(img_features.shape)
+            else:
+                img_features = self.rgb_encoder(
+                    einops.rearrange(batch["observation.images"], "b s n ... -> (b s n) ...")
+                )
+                # Separate batch dim and sequence dim back out. The camera index dim gets absorbed into the
+                # feature dim (effectively concatenating the camera features).
+                img_features = einops.rearrange(
+                    img_features, "(b s n) ... -> b s (n ...)", b=batch_size, s=n_obs_steps
+                )
+                print("using single rgb encoder")
+                print(img_features.shape)
             global_cond_feats.append(img_features)
 
         if self._use_env_state: