-
We don't see the latent code f ∈ R88 mentioned in paper 3.2. Is there any other data preprocessing step?
-
### 📚 The doc issue
Currently, the documentation for DeepLabV3 (segmentation) doesn't state both the minimum input size[^1] and the output stride of the backbone network, as configured by the origi…
-
is this code's first setting Second Network or Point Pillar Network?
If it is not Point Pillar Network, how can i change for that? and what is the difference?
-
Should `--prune_criterion l1-norm` be added in **./scripts/dist_train.sh**? I noticed that the default prune_criterion is act_scale: `parser.add_argument('--prune_criterion', type=str, default='act_sc…
-
After having a glance on channel pruning gpu version, i found that gpu version may not strictly implement lasso regression. Neither Coordinate descent method nor LARS optimization algorithm are used. …
-
My question:
In the file 'voxelnet.py': SparseMiddleExtractor
by the second "scn.Convolution", we can get the shape (64,2,400,352)
then in the "forward",by the "ret = ret.view(N, C * D…
-
Hi authors,
Thanks for your great implementation of sparse convolution and the paper~ I ran the 3D shape part segmentation experiments provided in `examples/3d_segmentation` recently and I modified…
-
It would be great if this new MAE-style method called SparK was introduced to lightly.
Paper: https://arxiv.org/abs/2301.03580 featured in ICLR'23 Spotlight
Code: https://github.com/keyu-tian/Spar…
-
## Issue description RuntimeError: Trying to create tensor with negative dimension -1741885395: [-1741885395]
I am running a graph convolution network model using Pytorch and get the problem:
```…
-
Facing the following issue while running a convolution network with sparse input layer.
> File: Matrix.cpp Line: 4659 Function: Microsoft::MSR::CNTK::Matrix::UnrollConvolutionInput -> Featu…