Merge pull request opencv#24672 from dkurt:adjust_slice_optional_inputs

Replace Slice optional inputs removal to adjustment
MengqingCao · Dec 10, 2023 · 098efb6 · 098efb6
2 parents 8b577ab + ac4b26a
commit 098efb6
Show file tree

Hide file tree

Showing 2 changed files with 26 additions and 36 deletions.
diff --git a/modules/dnn/src/onnx/onnx_graph_simplifier.cpp b/modules/dnn/src/onnx/onnx_graph_simplifier.cpp
@@ -182,12 +182,12 @@ static Mat extractConstant(const Ptr<ImportGraphWrapper>& net, int node_id, int
 }
 
 /*  Slice operator has two optional inputs "axes" and "steps". Some models may be set to have
-    Slice with optional inputs of default values, some of them don't. This Subgraph removes
-    all optional inputs of Slice if values are default.
+    Slice with optional inputs of default values, some of them don't. This Subgraph adjusts
+    all optional inputs of Slice up to 5.
 */
-class RemoveSliceAllOptionalInputsSubgraph : public Subgraph {
+class AdjustSliceAllOptionalInputsSubgraph : public Subgraph {
  public:
-    RemoveSliceAllOptionalInputsSubgraph(size_t num_inputs = 4) {
+    AdjustSliceAllOptionalInputsSubgraph(size_t num_inputs = 4) {
         num_inputs_ = num_inputs;
 
         int input = addNodeToMatch("");
@@ -200,35 +200,17 @@ class RemoveSliceAllOptionalInputsSubgraph : public Subgraph {
 
         slice_id = addNodeToMatch("Slice", inputs);
 
-        setFusedNode("Slice", std::vector<int>{input, starts, ends});
+        setFusedNode("Slice", inputs);
     }
 
-    virtual bool match(const Ptr<ImportGraphWrapper>& net, int nodeId,
-                       std::vector<int>& matchedNodesIds) CV_OVERRIDE {
-        if (Subgraph::match(net, nodeId, matchedNodesIds)) {
-            if (num_inputs_ >= 4) { // with axes
-                // Check if axes are -1 or last axis
-                auto onnx_net = net.dynamicCast<ONNXGraphWrapper>();
-                int shape_size = onnx_net->getTensorShapeSize(matchedNodesIds[slice_id], 0);
-
-                auto axes = extractConstant(net, matchedNodesIds[slice_id], 3);
-                for (size_t i = 0; i < axes.total(); i++) {
-                    const int axis = *(axes.ptr<const int>() + i);
-                    if (axis != -1 && axis != shape_size - 1) {
-                        return false;
-                    }
-                }
-            }
-            if (num_inputs_ == 5) {
-                // Check if steps are 1
-                auto steps = extractConstant(net, matchedNodesIds[slice_id], 4);
-                if (countNonZero(steps != 1)) {
-                    return false;
-                }
-            }
-            return true;
+    virtual void finalize(const Ptr<ImportGraphWrapper>&,
+                          const Ptr<ImportNodeWrapper>& fusedNode,
+                          std::vector<Ptr<ImportNodeWrapper> >&) CV_OVERRIDE
+    {
+        opencv_onnx::NodeProto* node = fusedNode.dynamicCast<ONNXNodeWrapper>()->node;
+        for (int i = num_inputs_; i < 5; ++i) {
+            node->add_input("");
         }
-        return false;
     }
 
 private:
@@ -1119,7 +1101,11 @@ class ResizeSubgraph1 : public ExtractScalesSubgraph
     ResizeSubgraph1() : ExtractScalesSubgraph()
     {
         int shape = addNodeToMatch("Shape", input);
-        int slice = addNodeToMatch("Slice", shape, addNodeToMatch("Constant"), addNodeToMatch("Constant"), addNodeToMatch("Constant"));
+        int slice = addNodeToMatch("Slice", {shape,
+                                             addNodeToMatch(""),
+                                             addNodeToMatch(""),
+                                             addNodeToMatch(""),
+                                             addNodeToMatch("")});
 
         int castConcat = addNodeToMatch("Cast", concatId);
         int concat = addNodeToMatch("Concat", slice, castConcat);
@@ -1163,7 +1149,11 @@ class ResizeSubgraph3 : public Subgraph
         int cast = addNodeToMatch("Cast", concat1);
 
         int shape2 = addNodeToMatch("Shape", input);
-        int slice = addNodeToMatch("Slice", shape2, addNodeToMatch("Constant"), addNodeToMatch("Constant"));
+        int slice = addNodeToMatch("Slice", {shape2,
+                                             addNodeToMatch(""),
+                                             addNodeToMatch(""),
+                                             addNodeToMatch(""),
+                                             addNodeToMatch("")});
         int concat2 = addNodeToMatch("Concat", slice, cast);
         addNodeToMatch("Resize", input, addNodeToMatch("Constant"), addNodeToMatch("Constant"), concat2);
 
@@ -1235,8 +1225,8 @@ class BatchNormalizationSubgraph2 : public BatchNormalizationSubgraphBase
 void simplifySubgraphs(opencv_onnx::GraphProto& net)
 {
     std::vector<Ptr<Subgraph> > subgraphs;
-    subgraphs.push_back(makePtr<RemoveSliceAllOptionalInputsSubgraph>(4));
-    subgraphs.push_back(makePtr<RemoveSliceAllOptionalInputsSubgraph>(5));
+    subgraphs.push_back(makePtr<AdjustSliceAllOptionalInputsSubgraph>(3));
+    subgraphs.push_back(makePtr<AdjustSliceAllOptionalInputsSubgraph>(4));
     subgraphs.push_back(makePtr<GeluSubGraph>());
     subgraphs.push_back(makePtr<GeluApproximationSubGraph>());
     subgraphs.push_back(makePtr<LayerNormSubGraph>());

diff --git a/modules/dnn/src/onnx/onnx_importer.cpp b/modules/dnn/src/onnx/onnx_importer.cpp
@@ -1235,7 +1235,7 @@ void ONNXImporter::parseSlice(LayerParams& layerParams, const opencv_onnx::NodeP
         starts_ = DictValue::arrayInt(start_blob.begin<int>(), start_blob.total());
         ends_ = DictValue::arrayInt(end_blob.begin<int>(), end_blob.total());
 
-        if (inp_size > 3)
+        if (inp_size > 3 && !getBlob(node_proto, 3).empty())
         {
             Mat axes_blob = getBlob(node_proto, 3);
             CV_Assert(axes_blob.total() == start_blob.total());
@@ -1244,7 +1244,7 @@ void ONNXImporter::parseSlice(LayerParams& layerParams, const opencv_onnx::NodeP
             has_axes = true;
         }
 
-        if (inp_size == 5)
+        if (inp_size == 5 && !getBlob(node_proto, 4).empty())
         {
             Mat step_blob = getBlob(node_proto, 4);
             CV_Assert(step_blob.total() == start_blob.total());