Use Case and High-Level Description

This is a TensorFlow* version of densenet-161 model, one of the DenseNet group of models designed to perform image classification. The weights were converted from DenseNet-Keras Models. For details see repository, paper.

Example

Specification

Metric	Value
Type	Classification
GFlops	14.128
MParams	28.666
Source framework	TensorFlow*

Accuracy

Metric	Value
Top 1	76.446%
Top 5	93.228%

Performance

Input

Original Model

Image, name: Placeholder , shape: [1x224x224x3], format: [BxHxWxC], where:

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

Expected color order: RGB. Mean values: [123.68, 116.78, 103.94], scale factor for each channel: 58.8235294

Converted Model

Image, name: Placeholder, shape: [1x3x224x224], format: [BxCxHxW], where:

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

Expected color order: BGR.

Output

Original Model

Floating point values in range [0, 1], which represent probabilities for classes in a dataset. Name: densenet161/predictions/Reshape_1.

Converted Model

Floating point values in a range [0, 1], which represent probabilities for classes in a dataset. Name: densenet161/predictions/Reshape_1/Transpose, shape: [1, 1, 1, 1000].

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-TF-DenseNet.txt.