diff --git a/inference-engine/src/mkldnn_plugin/mkldnn_plugin.cpp b/inference-engine/src/mkldnn_plugin/mkldnn_plugin.cpp
index e6afe674247..1c8b40aaa8b 100644
--- a/inference-engine/src/mkldnn_plugin/mkldnn_plugin.cpp
+++ b/inference-engine/src/mkldnn_plugin/mkldnn_plugin.cpp
@@ -43,6 +43,7 @@
 #include <transformations/op_conversions/convert_gelu.hpp>
 #include <transformations/op_conversions/convert_gather_downgrade.hpp>
 #include <transformations/op_conversions/convert_gather_upgrade.hpp>
+#include <transformations/op_conversions/detection_output_downgrade.hpp>
 #include <transformations/op_conversions/detection_output_upgrade.hpp>
 #include <transformations/op_conversions/gelu7_downgrade.hpp>
 #include <transformations/op_conversions/hswish_decomposition.hpp>
@@ -375,6 +376,7 @@ static void TransformationUpToCPUSpecificOpSet(std::shared_ptr<ngraph::Function>
     pass_config->disable<ngraph::pass::ConvertReduceMaxToPooling>();
     pass_config->disable<ngraph::pass::ConvertReduceSumToPooling>();
     pass_config->disable<ngraph::pass::SliceToStridedSlice>();
+    pass_config->disable<ngraph::pass::ConvertDetectionOutput8ToDetectionOutput1>();
 
     pass_config->enable<ngraph::pass::NormalizeL2Decomposition>();
     pass_config->enable<ngraph::pass::ConvertInterpolate1ToInterpolate4>();
diff --git a/src/common/transformations/src/transformations/common_optimizations/common_optimizations.cpp b/src/common/transformations/src/transformations/common_optimizations/common_optimizations.cpp
index da8a676f5bd..66e46db1ded 100644
--- a/src/common/transformations/src/transformations/common_optimizations/common_optimizations.cpp
+++ b/src/common/transformations/src/transformations/common_optimizations/common_optimizations.cpp
@@ -68,6 +68,7 @@
 #include "transformations/op_conversions/convert_gelu.hpp"
 #include "transformations/op_conversions/convert_interpolate1_to_interpolate4.hpp"
 #include "transformations/op_conversions/detection_output_downgrade.hpp"
+#include "transformations/op_conversions/detection_output_upgrade.hpp"
 #include "transformations/op_conversions/batch_norm_decomposition.hpp"
 #include "transformations/op_conversions/einsum_decomposition.hpp"
 #include "transformations/op_conversions/gelu7_downgrade.hpp"
@@ -178,6 +179,7 @@ bool ngraph::pass::CommonOptimizations::run_on_function(std::shared_ptr<ngraph::
     manager.register_pass<ngraph::pass::ConvertMaxPool8ToMaxPool1>();
     manager.register_pass<ngraph::pass::ConvertPriorBox8To0>();  // not plugins implemented priorbox8
     manager.register_pass<ngraph::pass::ConvertDetectionOutput8ToDetectionOutput1>();
+    manager.register_pass<ngraph::pass::ConvertDetectionOutput1ToDetectionOutput8, false>();
 
     auto fq_fusions = manager.register_pass<ngraph::pass::GraphRewrite>();
     fq_fusions->add_matcher<ngraph::pass::FakeQuantizeMulFusion>();