62 строки
2.6 KiB
Plaintext
62 строки
2.6 KiB
Plaintext
**Source Link**
|
|
|
|
https://github.com/DeepScale/SqueezeNet
|
|
|
|
**Download Link**
|
|
|
|
https://github.com/DeepScale/SqueezeNet/blob/master/SqueezeNet_v1.1/squeezenet_v1.1.caffemodel
|
|
https://github.com/DeepScale/SqueezeNet/blob/master/SqueezeNet_v1.1/train_val.prototxt
|
|
|
|
|
|
**Project Page**
|
|
|
|
https://github.com/DeepScale/SqueezeNet
|
|
|
|
**Authors**
|
|
|
|
Original Paper Title: SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size
|
|
Authors: Forrest N. Iandola and Song Han and Matthew W. Moskewicz and Khalid Ashraf and William J.
|
|
|
|
Caffe Implementation: http://deepscale.ai
|
|
|
|
**Labels**
|
|
|
|
Imagenet Labels
|
|
https://gist.github.com/yrevar/942d3a0ac09ec9e5eb3a
|
|
|
|
**License**
|
|
|
|
BSD License
|
|
|
|
Redistribution and use in source and binary forms, with or without modification, are permitted
|
|
provided that the following conditions are met:
|
|
|
|
1. Redistributions of source code must retain the above copyright notice, this list of conditions
|
|
and the following disclaimer.
|
|
|
|
2. Redistributions in binary form must reproduce the above copyright notice, this list of conditions
|
|
and the following disclaimer in the documentation and/or other materials provided with the
|
|
distribution.
|
|
|
|
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR
|
|
IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
|
|
FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR
|
|
CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
|
|
DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
|
|
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER
|
|
IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF
|
|
THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
|
|
|
https://github.com/DeepScale/SqueezeNet/blob/master/LICENSE
|
|
|
|
**Changes Required**
|
|
|
|
The train_valid.prototxt needs to be updated for deployment. To do so, the following changes are required:
|
|
|
|
1. The data layer should be replaced by an input layer with dimensions equal to the crop size in the data layer.
|
|
2. The accuracy layers from the end of the prototxt file need to be deleted as they are not required during
|
|
deployment. They can remain in the prototxt but they will be ignored.
|
|
3. The loss layer will be ignored, and a layer of type softmax is required to be added, with bottom layer as
|
|
pool10 and top as "prob"
|
|
4. In the conversion call, the biases provided in the train_valid data layer are required to be given as arguments
|