From 95e3a1bf5db9ef9fa07670b838dd91c8124e7ee9 Mon Sep 17 00:00:00 2001 From: Ilya Lavrenov Date: Wed, 10 Mar 2021 14:57:37 +0300 Subject: [PATCH] Ported #4657 (#4698) * Updated API changes document * Comment for CVS-49440 --- docs/IE_DG/API_Changes.md | 13 +++++++++++++ inference-engine/include/ie_core.hpp | 3 +++ 2 files changed, 16 insertions(+) diff --git a/docs/IE_DG/API_Changes.md b/docs/IE_DG/API_Changes.md index c23c427e6ed..f379d6d5790 100644 --- a/docs/IE_DG/API_Changes.md +++ b/docs/IE_DG/API_Changes.md @@ -2,6 +2,19 @@ The sections below contain detailed list of changes made to the Inference Engine API in recent releases. +## 2021.3 + +### New API + + * InferenceEngine::InferRequest::Cancel to cancel inference request execution + * InferenceEngine::Layout::HWC to support HWC layout for input or output blobs + * InferenceEngine::Precision::F64 data precision for f64 data type + * InferenceEngine::CNNNetwork::getOVNameForTensor to map frameworks tensor names to OpenVINO internal tensor names + +### Deprecated API + + * InferenceEngine::IVariableState interface is deprecated, use InferenceEngine::VariableState wrapper + ## 2021.2 ### New API diff --git a/inference-engine/include/ie_core.hpp b/inference-engine/include/ie_core.hpp index 41046a2087b..c5197f0a41c 100644 --- a/inference-engine/include/ie_core.hpp +++ b/inference-engine/include/ie_core.hpp @@ -85,6 +85,9 @@ public: * `InferenceEngine::Core::ReadNetwork(const std::string& model, const Blob::CPtr& weights) const` * function overload which takes a filesystem path to the model. * For ONNX case the second parameter should contain empty blob. + * @note Created InferenceEngine::CNNNetwork object shares the weights with `weights` object. + * So, do not create `weights` on temporary data which can be later freed, since the network + * constant datas become to point to invalid memory. * @return CNNNetwork */ CNNNetwork ReadNetwork(const std::string& model, const Blob::CPtr& weights) const;