pr

README.md

@@ -140,9 +140,12 @@ SparseConvNet is BSD licensed, as found in the LICENSE file. [Terms of use](http
 7. [Submanifold Sparse Convolutional Networks, 2017](https://arxiv.org/abs/1706.01307) Introduces deep 'submanifold' SparseConvNets.
 8. [Workshop on Learning to See from 3D Data, 2017](https://shapenet.cs.stanford.edu/iccv17workshop/) First place in the [semantic segmentation](https://shapenet.cs.stanford.edu/iccv17/) competition. [Report](https://arxiv.org/pdf/1710.06104)
 9. [3D Semantic Segmentation with Submanifold Sparse Convolutional Networks, 2017](https://arxiv.org/abs/1711.10275) Semantic segmentation for the ShapeNet Core55 and NYU-DepthV2 datasets, CVPR 2018
-10. [ScanNet 3D semantic label benchmark 2018](http://kaldir.vc.in.tum.de/scannet_benchmark/semantic_label_3d) 0.726 average IOU.
-11. [MinkowskiEngine](https://github.com/StanfordVL/MinkowskiEngine) is an alternative implementation of SparseConvNet.
-12. [OccuSeg](https://arxiv.org/abs/2003.06537) object detection using SparseConvNets
+10. [Unsupervised learning with sparse space-and-time autoencoders](https://arxiv.org/abs/1811.10355) (3+1)D space-time autoencoders
+11. [ScanNet 3D semantic label benchmark 2018](http://kaldir.vc.in.tum.de/scannet_benchmark/semantic_label_3d) 0.726 average IOU.
+12. [MinkowskiEngine](https://github.com/StanfordVL/MinkowskiEngine) is an alternative implementation of SparseConvNet; [0.736 average IOU for ScanNet]( https://github.com/chrischoy/SpatioTemporalSegmentation).
+13. [SpConv: PyTorch Spatially Sparse Convolution Library](https://github.com/traveller59/spconv) is an alternative implementation of SparseConvNet.
+14. [Live Semantic 3D Perception for Immersive Augmented Reality](https://ieeexplore.ieee.org/document/8998140) describes a way to optimize memory access for SparseConvNet.
+15. [OccuSeg](https://arxiv.org/abs/2003.06537) real-time object detection using SparseConvNets.
 
 ## Citations
 

setup.py

@@ -14,7 +14,7 @@
 this_dir = os.path.dirname(os.path.realpath(__file__))
 torch_dir = os.path.dirname(torch.__file__)
 conda_include_dir = '/'.join(torch_dir.split('/')[:-4]) + '/include'
-extra = {'cxx': ['-std=c++14', '-fopenmp'], 'nvcc': ['-std=c++14', '-Xcompiler', '-fopenmp']}
+extra = {'cxx': ['-std=c++14', '-fopenmp','-O3'], 'nvcc': ['-std=c++14', '-Xcompiler', '-fopenmp', '-O3']}
 
 setup(
     name='sparseconvnet',

sparseconvnet/averagePooling.py

@@ -29,7 +29,7 @@ def forward(self, input):
         output = SparseConvNetTensor()
         output.metadata = input.metadata
         output.spatial_size = (
-            input.spatial_size - self.pool_size) / self.pool_stride + 1
+            input.spatial_size - self.pool_size) // self.pool_stride + 1
         assert ((output.spatial_size - 1) * self.pool_stride +
                 self.pool_size == input.spatial_size).all()
         output.features = AveragePoolingFunction.apply(

sparseconvnet/convolution.py

@@ -33,7 +33,7 @@ def forward(self, input):
         output = SparseConvNetTensor()
         output.metadata = input.metadata
         output.spatial_size =\
-            (input.spatial_size - self.filter_size) / self.filter_stride + 1
+            (input.spatial_size - self.filter_size) // self.filter_stride + 1
         assert ((output.spatial_size - 1) * self.filter_stride +
                 self.filter_size == input.spatial_size).all(), (input.spatial_size,output.spatial_size,self.filter_size,self.filter_stride)
         output.features = ConvolutionFunction.apply(

sparseconvnet/maxPooling.py

@@ -80,7 +80,7 @@ def forward(self, input):
         output = SparseConvNetTensor()
         output.metadata = input.metadata
         output.spatial_size = (
-            input.spatial_size - self.pool_size) / self.pool_stride + 1
+            input.spatial_size - self.pool_size) // self.pool_stride + 1
         assert ((output.spatial_size - 1) * self.pool_stride +
                 self.pool_size == input.spatial_size).all()
         output.features = MaxPoolingFunction.apply(

sparseconvnet/randomizedStrideConvolution.py

@@ -44,7 +44,7 @@ def forward(self, input):
         output = SparseConvNetTensor()
         output.metadata = input.metadata
         output.spatial_size =\
-            (input.spatial_size - self.filter_size) / self.filter_stride + 1
+            (input.spatial_size - self.filter_size) // self.filter_stride + 1
         assert ((output.spatial_size - 1) * self.filter_stride +
                 self.filter_size == input.spatial_size).all()
         output.features = (RandomizedStrideConvolutionFunction if self.training else ConvolutionFunction).apply(

sparseconvnet/randomizedStrideMaxPooling.py

@@ -80,7 +80,7 @@ def forward(self, input):
         output = SparseConvNetTensor()
         output.metadata = input.metadata
         output.spatial_size = (
-            input.spatial_size - self.pool_size) / self.pool_stride + 1
+            input.spatial_size - self.pool_size) // self.pool_stride + 1
         assert ((output.spatial_size - 1) * self.pool_stride +
                 self.pool_size == input.spatial_size).all()
         output.features = (RandomizedStrideMaxPoolingFunction if self.training else MaxPoolingFunction).apply(