Removed redundency in RTDETRDecoder (#4118)

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
Co-authored-by: Laughing-q <1185102784@qq.com>
Co-authored-by: Laughing <61612323+Laughing-q@users.noreply.github.com>

ultralytics/nn/modules/head.py

@@ -307,8 +307,6 @@ class RTDETRDecoder(nn.Module):
         features = self.enc_output(valid_mask * feats)  # bs, h*w, 256
 
         enc_outputs_scores = self.enc_score_head(features)  # (bs, h*w, nc)
-        # dynamic anchors + static content
-        enc_outputs_bboxes = self.enc_bbox_head(features) + anchors  # (bs, h*w, 4)
 
         # query selection
         # (bs, num_queries)
@@ -316,22 +314,23 @@ class RTDETRDecoder(nn.Module):
         # (bs, num_queries)
         batch_ind = torch.arange(end=bs, dtype=topk_ind.dtype).unsqueeze(-1).repeat(1, self.num_queries).view(-1)
 
-        # Unsigmoided
+        # (bs, num_queries, 256)
-        refer_bbox = enc_outputs_bboxes[batch_ind, topk_ind].view(bs, self.num_queries, -1)
+        top_k_features = features[batch_ind, topk_ind].view(bs, self.num_queries, -1)
-        # refer_bbox = torch.gather(enc_outputs_bboxes, 1, topk_ind.reshape(bs, self.num_queries).unsqueeze(-1).repeat(1, 1, 4))
+        # (bs, num_queries, 4)
+        top_k_anchors = anchors[:, topk_ind].view(bs, self.num_queries, -1)
+
+        # dynamic anchors + static content
+        refer_bbox = self.enc_bbox_head(top_k_features) + top_k_anchors
 
         enc_bboxes = refer_bbox.sigmoid()
         if dn_bbox is not None:
             refer_bbox = torch.cat([dn_bbox, refer_bbox], 1)
-        if self.training:
-            refer_bbox = refer_bbox.detach()
         enc_scores = enc_outputs_scores[batch_ind, topk_ind].view(bs, self.num_queries, -1)
 
-        if self.learnt_init_query:
+        embeddings = self.tgt_embed.weight.unsqueeze(0).repeat(bs, 1, 1) if self.learnt_init_query else top_k_features
-            embeddings = self.tgt_embed.weight.unsqueeze(0).repeat(bs, 1, 1)
+        if self.training:
-        else:
+            refer_bbox = refer_bbox.detach()
-            embeddings = features[batch_ind, topk_ind].view(bs, self.num_queries, -1)
+            if not self.learnt_init_query:
-            if self.training:
                 embeddings = embeddings.detach()
         if dn_embed is not None:
             embeddings = torch.cat([dn_embed, embeddings], 1)