The octave-resnet-50-0.125
model is a modification of ResNet-50 from this paper with octave convolutions from Drop an Octave: Reducing Spatial Redundancy in Convolutional Neural Networks with Octave Convolution with alpha=0.125
. The model is originally designed for image classification. For details about family of Octave Convolution models, check out the repository.
The model input is a blob that consists of a single image of 1x3x224x224 in RGB order. The RGB mean values need to be subtracted as follows: [124,117,104] before passing the image blob into the network. In addition, values must be divided by 0.0167.
The model output for octave-resnet-50-0.125
is the typical object classifier output for the 1000 different classifications matching those in the ImageNet database.
Metric | Value |
---|---|
Type | Classification |
GFLOPs | 7.221 |
MParams | 25.551 |
Source framework | MXNet* |
Metric | Value |
---|---|
Top 1 | 78.19% |
Top 5 | 93.862% |
Image, name: data
, shape: 1,3,224,224
, format: B,C,H,W
, where:
B
- batch sizeC
- channelH
- heightW
- widthChannel order is RGB
. Mean values: [124,117,104], scale value: 59.880239521.
Image, name: data
, shape: 1,3,224,224
, format: B,C,H,W
,where:
B
- batch sizeC
- channelH
- heightW
- widthChannel order is BGR
Object classifier according to ImageNet classes, name: prob
, shape: 1,1000
, output data format is B,C
, where:
B
- batch sizeC
- predicted probabilities for each class in [0, 1] rangeObject classifier according to ImageNet classes, name: prob
, shape: 1,1000
, output data format is B,C
, where:
B
- batch sizeC
- predicted probabilities for each class in [0, 1] rangeThe original model is distributed under the following license: