dnlcrl / deep-residual-networks-pyfunt

Python implementation of "Deep Residual Learning for Image Recognition" (http://arxiv.org/abs/1512.03385 - MSRA, winner team of the 2015 ILSVRC and COCO challenges).

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Deep Residual Learning for Image Recognition

Implementation of "Deep Residual Learning for Image Recognition", Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun in PyFunt (a simple Python + Numpy DL framework).

Also inspired by this implementation in Lua + Torch.

The network operates on minibatches of data that have shape (N, C, H, W) consisting of N images, each with height H and width W and with C input channels. It has, like in the reference paper, (6*n)+2 layers, composed as below:

		                                        (image_dim: 3, 32, 32; F=16)
		                                        (input_dim: N, *image_dim)
		 INPUT
		    |
		    v
		+-------------------+
		|conv[F, *image_dim]|                    (out_shape: N, 16, 32, 32)
		+-------------------+
		    |
		    v
		+-------------------------+
		|n * res_block[F, F, 3, 3]|              (out_shape: N, 16, 32, 32)
		+-------------------------+
		    |
		    v
		+-------------------------+
		|res_block[2*F, F, 3, 3]  |              (out_shape: N, 32, 16, 16)
		+-------------------------+
		    |
		    v
		+---------------------------------+
		|(n-1) * res_block[2*F, 2*F, 3, 3]|      (out_shape: N, 32, 16, 16)
		+---------------------------------+
		    |
		    v
		+-------------------------+
		|res_block[4*F, 2*F, 3, 3]|              (out_shape: N, 64, 8, 8)
		+-------------------------+
		    |
		    v
		+---------------------------------+
		|(n-1) * res_block[4*F, 4*F, 3, 3]|      (out_shape: N, 64, 8, 8)
		+---------------------------------+
		    |
		    v
		+-------------+
		|pool[1, 8, 8]|                          (out_shape: N, 64, 1, 1)
		+-------------+
		    |
		    v
		+-------+
		|softmax|                                (out_shape: N, num_classes)
		+-------+
		    |
		    v
		 OUTPUT

Every convolution layer has a pad=1 and stride=1, except for the dimension enhancning layers which has a stride of 2 to mantain the computational complexity. Optionally, there is the possibility of setting m affine layers immediatley before the softmax layer by setting the hidden_dims parameter, which should be a list of integers representing the numbe of neurons for each affine layer.

Each residual block is composed as below:

          Input
             |
     ,-------+-----.
Downsampling      3x3 convolution+dimensionality reduction
    |               |
    v               v
Zero-padding      3x3 convolution
    |               |
    `-----( Add )---'
             |
          Output

After every layer, a batch normalization with momentum .1 is applied.

Requirements

After you get Python, you can get pip and install all requirements by running:

pip install -r requirements.txt

Usage

If you want to train the network on the CIFAR-10 dataset, simply run:

python train.py --help

Otherwise, you have to get the right train.py for MNIST or SFDDD datasets, they are respectively on the mnist and sfddd git branches:

Experiments Results

You can view all the experiments results in the ./docs directory. Main results are shown below:

best error: 9.59 % (accuracy: 0.9041) with a 20 layers residual network (n=3):

CIFAR-10 results

CIFAR-10 Results - iPython notebook

best error: 0.36 % (accuracy: 0.9964) with a 32 layers residual network (n=5):

MNIST results

MNIST Results - iPython notebook

best error: 0.25 % (accuracy: 0.9975 %) on a subset (1000 samples) of the train data (~21k images) with a 44 layers residual network (n=7), resizing the images to 64x48, randomly cropping 32x32 images for training and cropping a 32x32 image from the center of the original images for testing. Unfortunately I got more than 2% error on Kaggle's results (composed of ~80k images).

SFDDD results

SFDDD Results - iPython notebook

About

Python implementation of "Deep Residual Learning for Image Recognition" (http://arxiv.org/abs/1512.03385 - MSRA, winner team of the 2015 ILSVRC and COCO challenges).

License:MIT License


Languages

Language:Python 100.0%