Use Case and High-Level Description

This is a vehicle detection network based on an SSD framework with tuned MobileNet v1 as a feature extractor.



Metric Value
Average Precision (AP) 90.6%
Target vehicle size 40 x 30 pixels on Full HD image
Max objects to detect 200
GFlops 2.798
MParams 1.079
Source framework Caffe*

For Average Precision metric description, see The PASCAL Visual Object Classes (VOC) Challenge.

Tested on a challenging internal dataset with 3000 images and 12585 vehicles to detect.


Image, name: input, shape: 1, 3, 384, 672 in the format B, C, H, W, where:

  • B - batch size
  • C - number of channels
  • H - image height
  • W - image width

Expected color order is BGR.


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 (1 - vehicle)
  • 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

Legal Information

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