Unity导入自定义ONNX模型失败求助：Loop类型未识别错误

我在实现目标检测功能，参考YouTube教程用TensorFlow制作了模型，通过以下命令将SavedModel转为ONNX格式：

python -m tf2onnx.convert --saved-model TensorFlow-model-path --output model.onnx

但将该ONNX模型导入Unity时失败，导入他人的ONNX模型却能成功。报错核心是解析StatefulPartitionedCall/map/while_loop:0层时遇到未知类型Loop，具体错误日志如下：

错误日志1 #

OnnxImportException: Unknown type Loop encountered while parsing layer StatefulPartitionedCall/map/while_loop:0.

Unity.Barracuda.ONNX.ONNXModelConverter.Err (Unity.Barracuda.Model model, System.String layerName, System.String message, System.String extendedMessage, System.String debugMessage) (at Library/PackageCache/com.unity.barracuda@3.0.0/Barracuda/Runtime/ONNX/ONNXModelConverter.cs:3434)
Unity.Barracuda.ONNX.ONNXModelConverter.ConvertOnnxModel (Onnx.ModelProto onnxModel) (at Library/PackageCache/com.unity.barracuda@3.0.0/Barracuda/Runtime/ONNX/ONNXModelConverter.cs:2954)
Unity.Barracuda.ONNX.ONNXModelConverter.Convert (Google.Protobuf.CodedInputStream inputStream) (at Library/PackageCache/com.unity.barracuda@3.0.0/Barracuda/Runtime/ONNX/ONNXModelConverter.cs:170)
Unity.Barracuda.ONNX.ONNXModelConverter.Convert (System.String filePath) (at Library/PackageCache/com.unity.barracuda@3.0.0/Barracuda/Runtime/ONNX/ONNXModelConverter.cs:98)
Unity.Barracuda.ONNXModelImporter.OnImportAsset (UnityEditor.AssetImporters.AssetImportContext ctx) (at Library/PackageCache/com.unity.barracuda@3.0.0/Barracuda/Editor/ONNXModelImporter.cs:65)
UnityEditor.AssetImporters.ScriptedImporter.GenerateAssetData (UnityEditor.AssetImporters.AssetImportContext ctx) (at <d63e58d1bc4b489593159fa6218bc0e6>:0)
UnityEditorInternal.InternalEditorUtility:ProjectWindowDrag(HierarchyProperty, Boolean)
UnityEngine.GUIUtility:ProcessEvent(Int32, IntPtr, Boolean&)

错误日志2 #

Asset import failed, "Assets/Models/model.onnx" > OnnxImportException: Unknown type Loop encountered while parsing layer StatefulPartitionedCall/map/while_loop:0.

Unity.Barracuda.ONNX.ONNXModelConverter.Err (Unity.Barracuda.Model model, System.String layerName, System.String message, System.String extendedMessage, System.String debugMessage) (at Library/PackageCache/com.unity.barracuda@3.0.0/Barracuda/Runtime/ONNX/ONNXModelConverter.cs:3434)
Unity.Barracuda.ONNX.ONNXModelConverter.ConvertOnnxModel (Onnx.ModelProto onnxModel) (at Library/PackageCache/com.unity.barracuda@3.0.0/Barracuda/Runtime/ONNX/ONNXModelConverter.cs:2954)
Unity.Barracuda.ONNX.ONNXModelConverter.Convert (Google.Protobuf.CodedInputStream inputStream) (at Library/PackageCache/com.unity.barracuda@3.0.0/Barracuda/Runtime/ONNX/ONNXModelConverter.cs:170)
Unity.Barracuda.ONNX.ONNXModelConverter.Convert (System.String filePath) (at Library/PackageCache/com.unity.barracuda@3.0.0/Barracuda/Runtime/ONNX/ONNXModelConverter.cs:98)
Unity.Barracuda.ONNXModelImporter.OnImportAsset (UnityEditor.AssetImporters.AssetImportContext ctx) (at Library/PackageCache/com.unity.barracuda@3.0.0/Barracuda/Editor/ONNXModelImporter.cs:65)
UnityEditor.AssetImporters.ScriptedImporter.GenerateAssetData (UnityEditor.AssetImporters.AssetImportContext ctx) (at <d63e58d1bc4b489593159fa6218bc0e6>:0)
UnityEditorInternal.InternalEditorUtility:ProjectWindowDrag(HierarchyProperty, Boolean)
UnityEngine.GUIUtility:ProcessEvent(Int32, IntPtr, Boolean&)

UnityEngine.GUIUtility:ProcessEvent (int,intptr,bool&)

解决方案 #

• 升级Barracuda版本：你当前使用的是Barracuda 3.0.0，该版本对ONNX Loop算子支持有限，建议升级到最新稳定版，新版本会扩展算子兼容范围。
• 指定低版本ONNX opset转换：TensorFlow转ONNX时默认opset可能过高，修改转换命令指定较低的opset版本（如opset11或opset13）：

  python -m tf2onnx.convert --saved-model TensorFlow-model-path --output model.onnx --opset 13
  

• 固化模型形状：模型中的while_loop通常用于处理动态形状输出，导出TensorFlow模型时指定固定输入尺寸：
  • 导出SavedModel时添加input_signature参数明确输入张量形状；
  • 若使用预训练检测模型，选择静态图版本导出，替换动态循环节点。
• 简化ONNX模型：用ONNX Simplifier消除冗余节点后再导入Unity：

  pip install onnx-simplifier
  python -m onnxsim model.onnx simplified_model.onnx
  

• 更换模型导出路径：若上述方法无效，可尝试通过TensorFlow Lite中转：先转TFLite格式再转ONNX，或直接在Unity中使用TFLite模型（Unity官方提供TFLite插件支持）。

内容的提问来源于stack exchange，提问作者user19952742