yolo-v2-ava-sparse-70-0001

Use Case and High-Level Description

This is a re-implemented and re-trained version of YOLO v2 object detection network trained with VOC2012 training dataset. Network weight pruning is applied to sparsify convolution layers (70% of network parameters are set to zeros).

Example

Specification

Metric Value
Mean Average Precision (mAP) 62.9%
Flops 48.29Bn*
Source framework Tensorflow**

Average Precision metric described in: Mark Everingham et al. "The PASCAL Visual Object Classes (VOC) Challenge".

Tested on VOC 2012 validation dataset.

Performance

Inputs

  1. name: "input" , shape: [1x3x416x416] - An input image in the format [BxCxHxW], where:
    • B - batch size
    • C - number of channels
    • H - image height
    • W - image width. Expected color order is BGR.

Outputs

  1. The net outputs a blob with the shape: [1, 21125], which can be reshaped to [5, 25, 13, 13], where each number corresponds to [num_anchors, cls_reg_obj_params, y_loc, x_loc], respectively.
    • num_anchors: number of anchor boxes, each spatial location specified by y_loc and x_loc has 5 anchors.
    • cls_reg_obj_params: parameters for classification and regression. The values are made up of the followings:
      • Regression parameters (4)
      • Objectness score (1)
      • Class score (20)
    • y_loc and x_loc: spatial location of each grid.

Legal Information

[*] Same as the original implementation.

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