https://github.com/dneprdroid/tfsecured

Small library for TensorFlow proto model's (*.pb) encryption/decryption
https://github.com/dneprdroid/tfsecured

aes cpp decryption encryption golang ios machine-learning tensorflow tensorflow-proto-models

Last synced: 6 days ago
JSON representation

Small library for TensorFlow proto model's (*.pb) encryption/decryption

• Host: GitHub
• URL: https://github.com/dneprdroid/tfsecured
• Owner: dneprDroid
• License: mit
• Created: 2018-05-03T19:35:13.000Z (over 6 years ago)
• Default Branch: master
• Last Pushed: 2022-06-30T11:01:08.000Z (over 2 years ago)
• Last Synced: 2024-11-06T10:03:54.215Z (6 days ago)
• Topics: aes, cpp, decryption, encryption, golang, ios, machine-learning, tensorflow, tensorflow-proto-models
• Language: Objective-C++
• Homepage:
• Size: 33 MB
• Stars: 82
• Watchers: 5
• Forks: 21
• Open Issues: 3
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
## TFSecured

Small library for tensorflow proto models (*.pb) encryption/decryption.

## AES

You may use random string with random length like a key, then library calculates sha256 hash of it and uses as internal key with size 256 bits.

## Usage 

Copy sources from [TFSecured](/TFSecured) dir into your project.

 

C++ usage (see [TFPredictor.mm](/iosDemo/TFSecured-iOS/tf/predict/TFPredictor.mm)):

```cpp

    #include 

    ........

    tensorflow::GraphDef graph;

    // Decryption: 

    const std::string key = "JHEW8F7FE6F8E76W8F687WE6F8W8EF5";

    auto status = tfsecured::GraphDefDecryptAES(path,         // path to *.pb file (frozen graph)

                                                &graph,

                                                key);         // your key

    if (!status.ok()) {

        std::cout << status.error_message() << std::endl;

        return;

    }

    

    // Create session :

    std::unique_ptr session(NewSession(options));

    status = session->Create(graph);

    

    // Run session ....

```

Encrypt model via python script (see [encrypt_model.py](/python/encrypt_model.py)):

```bash

$ python encrypt_model.py   \

                           \  

                                          # optional, generated randomly by script 

```

## Dependencies

* TensorFlow

* OpenSSL

## iOS Demo (Digit recognizer)

Prepare repository (install Tensorflow via pods)

```bash

$ cd iosDemo

$ pod install 

``` 

And open it in Xcode.