mmdetection算法之DETR（0）

---

前言

简单介绍下mmdetection里的DETR算法，后续有时间会详细补充和修改。

DETR

基本网络结构
backbone:resnet50输出的最后一层
neck:无
head:transformer，编解码+FFN+最后的预测分类和回归

一、模型输出

我们直接看经过resnet50特征提取后的transformer部分

    def forward_single(self, x, img_metas):
        """"Forward function for a single feature level.

        Args:
            x (Tensor): Input feature from backbone's single stage, shape
                [bs, c, h, w].
            img_metas (list[dict]): List of image information.

        Returns:
            all_cls_scores (Tensor): Outputs from the classification head,
                shape [nb_dec, bs, num_query, cls_out_channels]. Note
                cls_out_channels should includes background.
            all_bbox_preds (Tensor): Sigmoid outputs from the regression
                head with normalized coordinate format (cx, cy, w, h).
                Shape [nb_dec, bs, num_query, 4].
        """
        # construct binary masks which used for the transformer.
        # NOTE following the official DETR repo, non-zero values representing
        # ignored positions, while zero values means valid positions.
        batch_size = x.size(0)
        input_img_h, input_img_w = img_metas[0]['batch_input_shape']
        masks = x.new_ones((batch_size, input_img_h, input_img_w)) # 建立个MASK
        for img_id in range(batch_size):    # [B,H,W]
            img_h, img_w, _ = img_metas[img_id]['img_shape']
            masks[img_id, :img_h, :img_w] = 0

        x = self.input_proj(x)  # 2048-256 [B,C,H,W]
        # interpolate masks to have the same spatial shape with x
        masks = F.interpolate(
            masks.unsqueeze(1), size=x.shape[-2:]).to(torch.bool).squeeze(1) # [B,C,H,W]
        # position encoding，下面有其代码
        pos_embed = self.positional_encoding(masks)  # [bs, embed_dim, h, w]
        # outs_dec: [nb_dec, bs, num_query, embed_dim] # 下面有代码
        outs_dec, _ = self.transformer(x, masks, self.query_embedding.weight,
                                       pos_embed)  # 返回nb_dec个

        all_cls_scores = self.fc_cls(outs_dec)
        all_bbox_preds = self.fc_reg(self.activate(
            self.reg_ffn(outs_dec))).sigmoid()
        return all_cls_scores, all_bbox_preds

位置编码

    def forward(self, mask):
        """Forward function for `SinePositionalEncoding`.

        Args:
            mask (Tensor): ByteTensor mask. Non-zero values representing
                ignored positions, while zero values means valid positions
                for this image. Shape [bs, h, w].

        Returns:
            pos (Tensor): Returned position embedding with shape
                [bs, num_feats*2, h, w].
        """
        # For convenience of exporting to ONNX, it's required to convert
        # `masks` from bool to int.
        mask = mask.to(torch.int)
        not_mask = 1 - mask  # logical_not
        y_embed = not_mask.cumsum(1, dtype=torch.float32)  # Y方向进行累加
        x_embed = not_mask.cumsum(2, dtype=torch.float32)  # Ｘ方向进行累加
        if self.normalize:  # 分别将两个方向上的累加结果归一化到[0,2N]区间
            y_embed = (y_embed + self.offset) / \
                      (y_embed[:, -1:, :] + self.eps) * self.scale
            x_embed = (x_embed + self.offset) / \
                      (x_embed[:, :, -1:] + self.eps) * self.scale
        dim_t = torch.arange(
            self.num_feats, dtype=torch.float32, device=mask.device)
        dim_t = self.temperature**(2 * (dim_t // 2) / self.num_feats)
        pos_x = x_embed[:, :, :, None] / dim_t
        pos_y = y_embed[:, :, :, None] / dim_t
        # use `view` instead of `flatten` for dynamically exporting to ONNX
        B, H, W = mask.size()
        pos_x = torch.stack(
            (pos_x[:, :, :, 0::2].sin(), pos_x[:, :, :, 1::2].cos()),
            dim=4).view(B, H, W, -1)
        pos_y = torch.stack(
            (pos_y[:, :, :, 0::2].sin(), pos_y[:, :, :, 1::2].cos()),
            dim=4).view(B, H, W, -1)
        pos = torch.cat((pos_y, pos_x), dim=3).permute(0, 3, 1, 2)
        return pos

总结

    这种检测模型的代码介绍，我现在就一部分一部分介绍，还没找到合适的编写方法，先这样吧，后面有更好的编写习惯和模式后，在整体进行修改补充下吧！