TorchSharpCatalog.ObjectDetection 方法
定义
重要
一些信息与预发行产品相关，相应产品在发行之前可能会进行重大修改。 对于此处提供的信息，Microsoft 不作任何明示或暗示的担保。
重载
ObjectDetection(MulticlassClassificationCatalog+MulticlassClassificationTrainers, ObjectDetectionTrainer+Options)
- Source:
- TorchSharpCatalog.cs
- Source:
- TorchSharpCatalog.cs
- Source:
- TorchSharpCatalog.cs
微调对象检测模型。
public static Microsoft.ML.TorchSharp.AutoFormerV2.ObjectDetectionTrainer ObjectDetection(this Microsoft.ML.MulticlassClassificationCatalog.MulticlassClassificationTrainers catalog, Microsoft.ML.TorchSharp.AutoFormerV2.ObjectDetectionTrainer.Options options);static member ObjectDetection : Microsoft.ML.MulticlassClassificationCatalog.MulticlassClassificationTrainers * Microsoft.ML.TorchSharp.AutoFormerV2.ObjectDetectionTrainer.Options -> Microsoft.ML.TorchSharp.AutoFormerV2.ObjectDetectionTrainer<Extension()>
Public Function ObjectDetection (catalog As MulticlassClassificationCatalog.MulticlassClassificationTrainers, options As ObjectDetectionTrainer.Options) As ObjectDetectionTrainer参数
- options
- ObjectDetectionTrainer.Options
完整的高级选项集。
返回
适用于
ObjectDetection(MulticlassClassificationCatalog+MulticlassClassificationTrainers, String, String, String, String, String, String, Int32)
- Source:
- TorchSharpCatalog.cs
- Source:
- TorchSharpCatalog.cs
- Source:
- TorchSharpCatalog.cs
微调对象检测模型。
public static Microsoft.ML.TorchSharp.AutoFormerV2.ObjectDetectionTrainer ObjectDetection(this Microsoft.ML.MulticlassClassificationCatalog.MulticlassClassificationTrainers catalog, string labelColumnName = "Label", string predictedLabelColumnName = "PredictedLabel", string scoreColumnName = "Score", string boundingBoxColumnName = "BoundingBoxes", string predictedBoundingBoxColumnName = "PredictedBoundingBoxes", string imageColumnName = "Image", int maxEpoch = 10);static member ObjectDetection : Microsoft.ML.MulticlassClassificationCatalog.MulticlassClassificationTrainers * string * string * string * string * string * string * int -> Microsoft.ML.TorchSharp.AutoFormerV2.ObjectDetectionTrainer<Extension()>
Public Function ObjectDetection (catalog As MulticlassClassificationCatalog.MulticlassClassificationTrainers, Optional labelColumnName As String = "Label", Optional predictedLabelColumnName As String = "PredictedLabel", Optional scoreColumnName As String = "Score", Optional boundingBoxColumnName As String = "BoundingBoxes", Optional predictedBoundingBoxColumnName As String = "PredictedBoundingBoxes", Optional imageColumnName As String = "Image", Optional maxEpoch As Integer = 10) As ObjectDetectionTrainer参数
- labelColumnName
- String
标签列名称。 应该是 keytype 的向量
- predictedLabelColumnName
- String
输出预测的标签列名称。 键类型的向量
- scoreColumnName
- String
输出分数列名称。 是浮点数的向量。
- boundingBoxColumnName
- String
边界框列名称。 是浮点数的向量。 值的顺序应为 x0 y0 x1 y1。
- predictedBoundingBoxColumnName
- String
输出边界框列名称。 是浮点数的向量。 值的顺序应为 x0 y0 x1 y1。
- imageColumnName
- String
包含图像数据的列名。 是 MLImage
- maxEpoch
- Int32
要运行多少个纪元。