LogS

> 最后一层的特征是分类和回归分支的输出。你可以得到倒数第二层的特征。 > > ```python > model.predict(..., embed=[22]) > ``` Thanks, it works now.

> 最后一层的特征是分类和回归分支的输出。你可以得到倒数第二层的特征。 > > ```python > model.predict(..., embed=[22]) > ```  想请问一下，这个[22]输出后是256，具体代表什么呢？这是倒数第二层，因为我输出print(results[0].masks.shape)为（3,640,640），3应该代表有三个物体，那么这里256具体后面是得到什么呢？想请您解释一下，谢谢

> Great to hear it's working for you! If you have any more questions, feel free to ask. May I ask what this [22] output of 256 specifically represents? This...

> The `[22]` output with a dimension of 256 represents the feature embeddings from the 22nd (penultimate) layer of the model. These 256 values correspond to the feature vector for...

> To obtain the segmentation features, the embedding number corresponds to the specific layer where segmentation features are extracted. Typically, this would be the dedicated segmentation head or a related...

> To access the penultimate layer's features in the segmentation model, you can specify the layer index during inference. If you're using `model.predict`, you can pass the `embed` argument with...

> @LogSSim to extract specific features using the YOLO segmentation model, identify the layer of interest by inspecting the model architecture (`print(model.model)`), then specify the desired layer index with the...

> @LogSSim to extract specific features using the YOLO segmentation model, identify the layer of interest by inspecting the model architecture (`print(model.model)`), then specify the desired layer index with the...

> 出现此问题的原因是`FeatureExtractor`类错误地对 进行了切片`model.model.children()`。YOLO 模型使用复合架构，其中的层不是简单的连续层。相反，您可以使用`forward_hooks`直接从特定层提取中间输出。这是一个示例方法： > > ```python > from ultralytics import YOLO > import torch > import torch.nn as nn > > # Load YOLO segmentation model > model =...

I want to set a waypoint, such as a camera coordinate. What could I do?