facenet-20180408-102900

Use Case and High-Level Description

FaceNet: A Unified Embedding for Face Recognition and Clustering. For details see the repository, paper

Specification

Metric

Value

Type

Face recognition

GFlops

2.846

MParams

23.469

Source framework

TensorFlow*

Accuracy

Metric

Value

LFW accuracy

99.14%

Input

Original model

  1. Image, name - batch_join:0, shape - 1, 160, 160, 3, format B, H, W, C, where:

    • B - batch size

    • H - image height

    • W - image width

    • C - number of channels

    Expected color order - RGB. Mean values - [127.5, 127.5, 127.5], scale factor for each channel - 128.0

  2. A boolean input, manages state of the graph (train/infer), name - phase_train, shape - 1.

Converted model

Image, name - image_batch/placeholder_port_0, shape - 1, 160, 160, 3, format B, H, W, C, where:

  • B - batch size

  • H - image height

  • W - image width

  • C - number of channels

Expected color order: BGR.

Output

Original model

Vector of floating-point values - face embeddings, Name - embeddings.

Converted model

Face embeddings, name - InceptionResnetV1/Bottleneck/BatchNorm/Reshape_1/Normalize, in format B,C, where:

  • B - batch size

  • C - row-vector of 512 floating-point values - face embeddings

Download a Model and Convert it into OpenVINO™ IR Format

You can download models and if necessary convert them into OpenVINO™ IR format using the Model Downloader and other automation tools as shown in the examples below.

An example of using the Model Downloader:

omz_downloader --name <model_name>

An example of using the Model Converter:

omz_converter --name <model_name>

Demo usage

The model can be used in the following demos provided by the Open Model Zoo to show its capabilities: