Some question about Spatial Aggregation and Spatital-wise Temporal Causal Self-Attention

[pisces365]

Hi, there. Thank you for your wonderful work. I have two questions to ask.

1. Where is the Spatial Aggregation mentioned in the paper reflected in the code?
2. Does the Spatital-wise Temporal Causal Self-Attention refer to the following code block?

#   PlanUtransformer.py    line 327
            for pose_temporal_attn, pose_temporal_norm, spatial_attn, spatial_norm, ffn, ffn_norm in pose_attn_en:
                # 自注意力、归一化
                pose_queries = pose_queries + pose_temporal_attn(pose_queries, pose_tokens, pose_tokens, need_weights=False, attn_mask=self.attn_mask)[0]
                pose_queries = pose_temporal_norm(pose_queries)
                #b, f, h, w, c = queries.shape
                pose_queries = rearrange(pose_queries, 'b f c -> (b f) 1 c')
                queries = rearrange(queries, 'b f h w c -> (b f) (h w) c')
                # 空间注意力层、 归一化
                pose_queries = pose_queries + spatial_attn(pose_queries, queries, queries, need_weights=False, attn_mask=None)[0]
                pose_queries = spatial_norm(pose_queries)

                # 前馈网络，对空间注意力的结果进行进一步的非线性变换，增强模型的表达能力  对前馈网络的输出进行归一化。
                pose_queries = pose_queries + ffn(pose_queries)
                pose_queries = ffn_norm(pose_queries)
                pose_queries = rearrange(pose_queries, '(b f) 1 c -> b f c', b=b, f=f)
                queries = rearrange(queries, '(b f) (h w) c -> b f h w c', b=b, f=f, h=h, w=w)

#   PlanUtransformer.py    line 425
            for pose_temporal_attn, pose_temporal_norm, spatial_attn, spatial_norm, ffn, ffn_norm in pose_attn_de:
                # 自注意力、归一化
                pose_queries = pose_queries + pose_temporal_attn(pose_queries, pose_tokens, pose_tokens, need_weights=False, attn_mask=self.attn_mask)[0]
                pose_queries = pose_temporal_norm(pose_queries)
                #b, f, h, w, c = queries.shape
                pose_queries = rearrange(pose_queries, 'b f c -> (b f) 1 c')
                #queries = rearrange(queries, 'b f h w c -> (b f) (h w) c')
                queries = rearrange(queries, '(b f) c h w -> (b f) (h w) c', b=b, f=f, h=h, w=w)
                # 空间注意力、归一化
                pose_queries = pose_queries + spatial_attn(pose_queries, queries, queries, need_weights=False, attn_mask=None)[0]
                pose_queries = spatial_norm(pose_queries)
                
                pose_queries = pose_queries + ffn(pose_queries)
                pose_queries = ffn_norm(pose_queries)
                queries = rearrange(queries, '(b f) (h w) c -> (b f) c h w', b=b, f=f, h=h, w=w)
                pose_queries = rearrange(pose_queries, '(b f) 1 c -> b f c', b=b, f=f)
            pose_queries = pose_de_(pose_queries)
            pose_tokens = pose_de_(pose_tokens)