- An optimized PyTorch package with CUDA backend, including Encoding Layer :class:`encoding.nn.Encoding`, Multi-GPU Synchronized Batch Normalization :class:`encoding.nn.BatchNorm2d` and other customized modules and functions.
An optimized PyTorch package with CUDA backend.
- **Example Systems** for Semantic Segmentation (coming), CIFAR-10 Classification, `Texture Recognition <experiments/texture.html>`_ and `Style Transfer <experiments/style.html>`_ are provided in experiments section.
.. todo::
A PyTorch DataParallel compatible Synchronized Cross-GPU Batch Normalization will be provided soon.
.. toctree::
.. toctree::
:glob:
:glob:
...
@@ -19,6 +19,13 @@ Created by `Hang Zhang <http://hangzh.com/>`_
...
@@ -19,6 +19,13 @@ Created by `Hang Zhang <http://hangzh.com/>`_
notes/*
notes/*
.. toctree::
:glob:
:maxdepth: 1
:caption: Experiment Systems
experiments/*
.. toctree::
.. toctree::
:maxdepth: 1
:maxdepth: 1
:caption: Package Reference
:caption: Package Reference
...
@@ -31,13 +38,6 @@ Created by `Hang Zhang <http://hangzh.com/>`_
...
@@ -31,13 +38,6 @@ Created by `Hang Zhang <http://hangzh.com/>`_
Aggregate operation, aggregate the residuals of inputs (:math:`X`) with repect to the codewords (:math:`C`) with assignment weights (:math:`A`).
.. math::
e_{k} = \sum_{i=1}^{N} a_{ik} (x_i - d_k)
Shape:
- Input: :math:`A\in\mathcal{R}^{B\times N\times K}` :math:`X\in\mathcal{R}^{B\times N\times D}` :math:`C\in\mathcal{R}^{K\times D}` (where :math:`B` is batch, :math:`N` is total number of features, :math:`K` is number is codewords, :math:`D` is feature dimensions.)
- Input: :math:`X\in\mathcal{R}^{B\times N\times D}` :math:`C\in\mathcal{R}^{K\times D}` :math:`S\in \mathcal{R}^K` (where :math:`B` is batch, :math:`N` is total number of features, :math:`K` is number is codewords, :math:`D` is feature dimensions.)
- Input: :math:`X\in\mathcal{R}^{B\times N\times D}` :math:`C\in\mathcal{R}^{K\times D}` (where :math:`B` is batch, :math:`N` is total number of features, :math:`K` is number is codewords, :math:`D` is feature dimensions.)
- Input: :math:`R\in\mathcal{R}^{B\times N\times K\times D}` :math:`S\in \mathcal{R}^K` (where :math:`B` is batch, :math:`N` is total number of features, :math:`K` is number is codewords, :math:`D` is feature dimensions.)
- Input: :math:`X\in\mathcal{R}^{B\times N\times D}` or :math:`\mathcal{R}^{B\times D\times H\times W}` (where :math:`B` is batch, :math:`N` is total number of features or :math:`H\times W`.)
Hang Zhang, Jia Xue, and Kristin Dana. "Deep TEN: Texture Encoding Network." *The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2017*
Aggregate operation, aggregate the residuals (:math:`R`) with
assignment weights (:math:`A`).
.. math::
e_{k} = \sum_{i=1}^{N} a_{ik} r_{ik}
Shape:
- Input: :math:`A\in\mathcal{R}^{B\times N\times K}` :math:`R\in\mathcal{R}^{B\times N\times K\times D}` (where :math:`B` is batch, :math:`N` is total number of features, :math:`K` is number is codewords, :math:`D` is feature dimensions.)
