[Examples] Add the wasmedge-tensorflow plug-in examples.

Signed-off-by: YiYing He <[email protected]>

README.md

@@ -12,36 +12,13 @@
 
 ### Introduction
 
-This project provides the examples of high-level [wasi-nn] bindings on Rust programming language. Developers can refer to this project to write their machine learning application in a high-level language using the bindings, compile it to WebAssembly, and run it with a WebAssembly runtime that supports the [wasi-nn] proposal, such as [WasmEdge].
-
-> __NOTE__: These bindings are experimental (use at your own risk) and subject to upstream changes in the [wasi-nn] specification.
+This project provides the examples of high-level [wasi-nn] bindings and WasmEdge-TensorFlow plug-ins on Rust programming language. Developers can refer to this project to write their machine learning application in a high-level language using the bindings, compile it to WebAssembly, and run it with a WebAssembly runtime that supports the [wasi-nn] proposal, such as [WasmEdge].
 
 ### Prerequisites
 
-#### OpenVINO Installation
-
-Developers should install the [OpenVINO] first before build and run WasmEdge with wasi-nn and the examples.
-For this project, we use the version `2021.4.582`. Please refer to the [installation script](scripts/install_openvino.sh).
-
-```bash
-curl -sSL https://apt.repos.intel.com/openvino/2021/GPG-PUB-KEY-INTEL-OPENVINO-2021 | sudo gpg --dearmor > /usr/share/keyrings/GPG-PUB-KEY-INTEL-OPENVINO-2021.gpg
-echo "deb [signed-by=/usr/share/keyrings/GPG-PUB-KEY-INTEL-OPENVINO-2021.gpg] https://apt.repos.intel.com/openvino/2021 all main" | sudo tee /etc/apt/sources.list.d/intel-openvino-2021.list
-sudo apt update
-sudo apt install -y intel-openvino-runtime-ubuntu20-2021.4.582
-```
-
-After the installation, you may need to add the path:
-
-```bash
-source /opt/intel/openvino_2021/bin/setupvars.sh
-sudo ldconfig
-```
-
-[OpenVINO]: https://www.intel.com/content/www/us/en/developer/tools/openvino-toolkit/overview.html
-
 #### Rust Installation
 
-Please refer to the [Rust Official Site](https://www.rust-lang.org/tools/install) for the Rust installation.
+For building the WASM files from rust source, please refer to the [Rust Official Site](https://www.rust-lang.org/tools/install) for the Rust installation.
 After the installation, developers should add the `wasm32-wasi` target.
 
 ```bash
@@ -50,32 +27,22 @@ rustup target add wasm32-wasi
 
 #### Download the `wasi-nn` Rust Crate
 
-In Rust, download the [crate from crates.io][crates.io] by adding `wasi-nn = "0.1"` as a Cargo dependency.
-
-[crates.io]: https://crates.io/crates/wasi-nn
+In Rust, download the [crate from crates.io](https://crates.io/crates/wasi-nn) by adding `wasi-nn = "0.4.0"` as a Cargo dependency.
 
-#### Build the WasmEdge with WASI-NN supporting
+For using WasmEdge-TensorFlow plug-ins, please download the [crate from crates.io](https://crates.io/crates/wasmedge_tensorflow_interface) by adding `wasmedge_tensorflow_interface = "0.3.0"` as a Cargo dependency.
 
-For running the examples, developers should [build WasmEdge from source](https://wasmedge.org/book/en/extend/build.html).
-First developers should get the source:
+#### WasmEdge Installation
 
-```bash
-git clone https://github.com/WasmEdge/WasmEdge.git
-cd WasmEdge
-```
-
-And build with the `WASMEDGE_PLUGIN_WASI_NN_BACKEND` argument:
-
-```bash
-cmake -Bbuild -GNinja -WASMEDGE_PLUGIN_WASI_NN_BACKEND="OpenVINO" .
-cmake --build build
-# For the WASI-NN plugin, you should install this project.
-cmake --install build
-```
+You can refer to [here to install WasmEdge](https://wasmedge.org/docs/start/install#install).
 
-After the installation, developers can execute the `wasmedge` executable with WASI-NN plugin.
+For the examples with different wasi-nn backends or using the WasmEdge-Tensorflow plug-ins, please install with plug-ins and their dependencies:
 
-> Notice: If you didn't install the project, you should give the `WASMEDGE_PLUGIN_PATH` environment variable for specifying the WASI-NN plugin path (the built plugin is at `build/plugins/wasi_nn`).
+- [wasi-nn plug-in with OpenVINO backend](https://wasmedge.org/docs/start/install#wasi-nn-plug-in-with-openvino-backend)
+- [wasi-nn plug-in with PyTorch backend](https://wasmedge.org/docs/start/install#wasi-nn-plug-in-with-pytorch-backend)
+- [wasi-nn plug-in with PyTorch backend](https://wasmedge.org/docs/start/install#wasi-nn-plug-in-with-tensorflow-lite-backend)
+- [WasmEdge-Image plug-in](https://wasmedge.org/docs/start/install#wasmedge-image-plug-in)
+- [WasmEdge-TensorFlow plug-in](https://wasmedge.org/docs/start/install#wasmedge-tensorflow-plug-in)
+- [WasmEdge-TensorFlow-Lite plug-in](https://wasmedge.org/docs/start/install#wasmedge-tensorflow-lite-plug-in)
 
 ### Examples
 
@@ -86,6 +53,7 @@ After the installation, developers can execute the `wasmedge` executable with WA
 - [WASI]
 - [wasi-nn]
 - [WasmEdge]
+- [WasmEdge-TensorFlow rust interface](https://crates.io/crates/wasmedge_tensorflow_interface)
 - [wasi-nn-guest](https://github.com/radu-matei/wasi-nn-guest)
 
 [WasmEdge]: https://wasmedge.org/

pytorch-mobilenet-image/README.md

@@ -56,7 +56,7 @@ python3 gen_tensor input.jpg image-1x3x224x224.rgb
 
 ### Execute
 
-Users should [install the WasmEdge with WASI-NN PyTorch backend plug-in](https://wasmedge.org/book/en/write_wasm/rust/wasinn.html#get-wasmedge-with-wasi-nn-plug-in-pytorch-backend).
+Users should [install the WasmEdge with WASI-NN PyTorch backend plug-in](https://wasmedge.org/docs/start/install#wasi-nn-plug-in-with-pytorch-backend).
 
 Execute the WASM with the `wasmedge` with PyTorch supporting:
 

tflite-birds_v1-image/README.md

@@ -40,7 +40,7 @@ The `tflite` model is located at `./lite-model_aiy_vision_classifier_birds_V1_3.
 
 ### Execute
 
-Users should [install the WasmEdge with WASI-NN TensorFlow-Lite backend plug-in](https://wasmedge.org/book/en/write_wasm/rust/wasinn.html#get-wasmedge-with-wasi-nn-plug-in-tensorflow-lite-backend).
+Users should [install the WasmEdge with WASI-NN TensorFlow-Lite backend plug-in](https://wasmedge.org/docs/start/install#wasi-nn-plug-in-with-tensorflow-lite-backend).
 
 Execute the WASM with the `wasmedge` with Tensorflow Lite supporting:
 

wasmedge-tf-mobilenet_v2/README.md

@@ -0,0 +1,59 @@
+# Mobilenet Example For WasmEdge-Tensorflow plug-in
+
+This package is a high-level Rust bindings for [WasmEdge-TensorFlow plug-in](https://wasmedge.org/docs/develop/rust/tensorflow) example of Mobilenet.
+
+## Dependencies
+
+This crate depends on the `wasmedge_tensorflow_interface` in the `Cargo.toml`:
+
+```toml
+[dependencies]
+wasmedge_tensorflow_interface = "0.3.0"
+```
+
+## Build
+
+Compile the application to WebAssembly:
+
+```bash
+cd rust && cargo build --target=wasm32-wasi --release
+```
+
+The output WASM file will be at [`rust/target/wasm32-wasi/release/wasmedge-tf-example-mobilenet.wasm`](wasmedge-tf-example-mobilenet.wasm).
+To speed up the image processing, we can enable the AOT mode in WasmEdge with:
+
+```bash
+wasmedge compile rust/target/wasm32-wasi/release/wasmedge-tf-example-mobilenet.wasm wasmedge-tf-example-mobilenet_aot.wasm
+```
+
+## Run
+
+### Test data
+
+The testing image is located at `./grace_hopper.jpg`:
+
+![grace_hopper](grace_hopper.jpg)
+
+The frozen `pb` model is located at `./mobilenet_v2_1.4_224_frozen.pb`.
+
+### Execute
+
+Users should [install the WasmEdge with WasmEdge-TensorFlow and WasmEdge-Image plug-ins](https://wasmedge.org/docs/start/install#wasmedge-tensorflow-plug-in).
+
+```bash
+curl -sSf https://raw.githubusercontent.com/WasmEdge/WasmEdge/master/utils/install.sh | bash -s -- --plugins wasmedge_tensorflow wasmedge_image
+```
+
+Execute the WASM with the `wasmedge` with Tensorflow Lite supporting:
+
+```bash
+wasmedge --dir .:. wasmedge-tf-example-mobilenet.wasm mobilenet_v2_1.4_224_frozen.pb grace_hopper.jpg
+```
+
+You will get the output:
+
+```console
+653 : 0.3230184316635132
+```
+
+Which is index 653 (0-based index) with rate `0.33690106868743896`. The index 653 of label table (which is line 654 in `imagenet_slim_labels.txt`) is `military uniform`.