pytorch-semseg 0.1.2

pytorch-semseg
==============

|license|

Semantic Segmentation Algorithms Implemented in PyTorch
-------------------------------------------------------

This repository aims at mirroring popular semantic segmentation
architectures in PyTorch.

.. raw:: html

<p align="center">

.. raw:: html

</p>

Networks implemented
~~~~~~~~~~~~~~~~~~~~

- `PSPNet <https://arxiv.org/abs/1612.01105>`__ - With support for
loading pretrained models w/o caffe dependency
- `FRRN <https://arxiv.org/abs/1611.08323>`__ - Model A and B
- `FCN <https://arxiv.org/abs/1411.4038>`__ - All 1 (FCN32s), 2
(FCN16s) and 3 (FCN8s) stream variants
- `U-Net <https://arxiv.org/abs/1505.04597>`__ - With optional
deconvolution and batchnorm
- `Link-Net <https://codeac29.github.io/projects/linknet/>`__ - With
multiple resnet backends
- `Segnet <https://arxiv.org/abs/1511.00561>`__ - With Unpooling using
Maxpool indices

Upcoming
^^^^^^^^

- `E-Net <https://arxiv.org/abs/1606.02147>`__
- `RefineNet <https://arxiv.org/abs/1611.06612>`__

DataLoaders implemented
~~~~~~~~~~~~~~~~~~~~~~~

- `CamVid <http://mi.eng.cam.ac.uk/research/projects/VideoRec/CamVid/>`__
- `Pascal
VOC <http://host.robots.ox.ac.uk/pascal/VOC/voc2012/segexamples/index.html>`__
- `ADE20K <http://groups.csail.mit.edu/vision/datasets/ADE20K/>`__
- `MIT Scene Parsing
Benchmark <http://data.csail.mit.edu/places/ADEchallenge/ADEChallengeData2016.zip>`__
- `Cityscapes <https://www.cityscapes-dataset.com/>`__

Upcoming
^^^^^^^^

- `NYUDv2 <http://cs.nyu.edu/~silberman/datasets/nyu_depth_v2.html>`__
- `Sun-RGBD <http://rgbd.cs.princeton.edu/>`__
- `MS COCO <http://mscoco.org/>`__

Requirements
~~~~~~~~~~~~

- pytorch >=0.3.0
- torchvision ==0.2.0
- visdom >=1.0.1 (for loss and results visualization)
- scipy
- tqdm

One-line installation
^^^^^^^^^^^^^^^^^^^^^

``pip install -r requirements.txt``

Data
~~~~

- Download data for desired dataset(s) from list of URLs
`here <https://meetshah1995.github.io/semantic-segmentation/deep-learning/pytorch/visdom/2017/06/01/semantic-segmentation-over-the-years.html#sec_datasets>`__.
- Extract the zip / tar and modify the path appropriately in
``config.json``

Usage
~~~~~

Launch `visdom <https://github.com/facebookresearch/visdom#launch>`__ by
running (in a separate terminal window)

::

python -m visdom.server

**To train the model :**

::

python train.py [-h] [--arch [ARCH]] [--dataset [DATASET]]
[--img_rows [IMG_ROWS]] [--img_cols [IMG_COLS]]
[--n_epoch [N_EPOCH]] [--batch_size [BATCH_SIZE]]
[--l_rate [L_RATE]] [--feature_scale [FEATURE_SCALE]]
[--visdom [VISDOM]]

--arch Architecture to use ['fcn8s, unet, segnet etc']
--dataset Dataset to use ['pascal, camvid, ade20k etc']
--img_rows Height of the input image
--img_cols Width of the input image
--n_epoch # of the epochs
--batch_size Batch Size
--l_rate Learning Rate
--feature_scale Divider for # of features to use
--visdom Show visualization(s) on visdom | False by default

**To validate the model :**

::

python validate.py [-h] [--model_path [MODEL_PATH]] [--dataset [DATASET]]
[--img_rows [IMG_ROWS]] [--img_cols [IMG_COLS]]
[--batch_size [BATCH_SIZE]] [--split [SPLIT]]

--model_path Path to the saved model
--dataset Dataset to use ['pascal, camvid, ade20k etc']
--img_rows Height of the input image
--img_cols Width of the input image
--batch_size Batch Size
--split Split of dataset to validate on

**To test the model w.r.t. a dataset on custom images(s):**

::

python test.py [-h] [--model_path [MODEL_PATH]] [--dataset [DATASET]]
[--dcrf [DCRF]] [--img_path [IMG_PATH]] [--out_path [OUT_PATH]]

--model_path Path to the saved model
--dataset Dataset to use ['pascal, camvid, ade20k etc']
--dcrf Enable DenseCRF based post-processing
--img_path Path of the input image
--out_path Path of the output segmap

.. |license| image:: https://img.shields.io/github/license/mashape/apistatus.svg
:target: https://github.com/meetshah1995/pytorch-semseg/blob/master/LICENSE