Use Case and High-Level Description

The ssd_mobilenet_v2_coco model is a Single-Shot multibox Detection (SSD) network intended to perform object detection. The model has been trained from the Common Objects in Context (COCO) image dataset.

The model input is a blob that consists of a single image of 1x3x300x300 in RGB order.

The model output is a typical vector containing the tracked object data, as previously described. Note that the "class_id" data is now significant and should be used to determine the classification for any detected object.



Metric Value
Type Detection
GFLOPs 3.775
MParams 16.818
Source framework TensorFlow*


Metric Value
coco_precision 24.9452%



Note that original model expects image in RGB format, converted model - in BGR format.

Original model

Image, shape - 1,300,300,3, format is B,H,W,C where:

Channel order is RGB.

Converted model

Image, name - image_tensor, shape - 1,300,300,3, format is B,H,W,C where:

Channel order is BGR.


NOTE output format changes after Model Optimizer conversion. To find detailed explanation of changes, go to Model Optimizer development guide

Original model

  1. Classifier, name - detection_classes, contains predicted bounding boxes classes in range [1, 91]. The model was trained on Microsoft* COCO dataset version with 90 categories of object.
  2. Probability, name - detection_scores, contains probability of detected bounding boxes.
  3. Detection box, name - detection_boxes, contains detection boxes coordinates in format [y_min, x_min, y_max, x_max], where (x_min, y_min) are coordinates top left corner, (x_max, y_max) are coordinates right bottom corner. Coordinates are rescaled to input image size.
  4. Detections number, name - num_detections, contains the number of predicted detection boxes.

Converted model

The array of summary detection information, name - detection_out, shape - 1, 1, N, 7, where N is the number of detected bounding boxes. For each detection, the description has the format: [image_id, label, conf, x_min, y_min, x_max, y_max], where:

Legal Information

The original model is distributed under the Apache License, Version 2.0. A copy of the license is provided in APACHE-2.0-TensorFlow.txt.