face-detection-0202

Use Case and High-Level Description

Face detector based on MobileNetV2 as a backbone with a multiple SSD head for indoor and outdoor scenes shot by a front-facing camera. During the training of this model, training images were resized to 384x384.

Example

_images/face-detection-0202.png

Specification

Metric

Value

AP ( WIDER )

91.94%

GFlops

1.767

MParams

1.842

Source framework

PyTorch*

Average Precision (AP) is defined as an area under the precision/recall curve. All numbers were evaluated by taking into account only faces bigger than 64 x 64 pixels.

Inputs

Image, name: input, shape: 1, 3, 384, 384 in the format B, C, H, W, where:

  • B - batch size

  • C - number of channels

  • H - image height

  • W - image width

Expected color order: BGR.

Outputs

The net outputs blob with shape: 1, 1, 200, 7 in the format 1, 1, N, 7, where N is the number of detected bounding boxes. Each detection has the format [image_id, label, conf, x_min, y_min, x_max, y_max], where:

  • image_id - ID of the image in the batch

  • label - predicted class ID (0 - face)

  • conf - confidence for the predicted class

  • (x_min, y_min) - coordinates of the top left bounding box corner

  • (x_max, y_max) - coordinates of the bottom right bounding box corner

Training Pipeline

The OpenVINO Training Extensions provide a training pipeline, allowing to fine-tune the model on custom dataset.

Use Case and High-Level Description

Face detector based on MobileNetV2 as a backbone with a multiple SSD head for indoor and outdoor scenes shot by a front-facing camera. During the training of this model, training images were resized to 384x384.

Example

_images/face-detection-0202.png

Specification

Metric

Value

AP ( WIDER )

91.94%

GFlops

1.767

MParams

1.842

Source framework

PyTorch*

Average Precision (AP) is defined as an area under the precision/recall curve. All numbers were evaluated by taking into account only faces bigger than 64 x 64 pixels.

Inputs

Image, name: input, shape: 1, 3, 384, 384 in the format B, C, H, W, where:

  • B - batch size

  • C - number of channels

  • H - image height

  • W - image width

Expected color order: BGR.

Outputs

The net outputs blob with shape: 1, 1, 200, 7 in the format 1, 1, N, 7, where N is the number of detected bounding boxes. Each detection has the format [image_id, label, conf, x_min, y_min, x_max, y_max], where:

  • image_id - ID of the image in the batch

  • label - predicted class ID (0 - face)

  • conf - confidence for the predicted class

  • (x_min, y_min) - coordinates of the top left bounding box corner

  • (x_max, y_max) - coordinates of the bottom right bounding box corner

Training Pipeline

The OpenVINO Training Extensions provide a training pipeline, allowing to fine-tune the model on custom dataset.

Legal Information

[*] Other names and brands may be claimed as the property of others.