Hello,
I’m new to pytorch3d rasterization. I wanted to plot facial landmarks with shape while training my model So I came across a GitHub which is doing similar to that and using the same face model.
But now I am getting below issue. I tried many things but the issue still persists. Any suggestion where should I look in my code.
File “/home/kak/Documents/Thesis/Reconstruction/train_copy.py”, line 125, in train_epoch
** shape_images = self.render.render_shape(pred_vert, trans_vertices, input_image)**
** File “/home/kak/Documents/Thesis/Reconstruction/renderer.py”, line 274, in render_shape**
** rendering = self.rasterizer(transformed_vertices, self.faces.expand(batch_size, -1, -1), attributes)**
** File “/home/kak/anaconda3/envs/gt/lib/python3.6/site-packages/torch/nn/modules/module.py”, line 722, in _call_impl**
** result = self.forward(input, kwargs)
** File “/home/kak/Documents/Thesis/Reconstruction/renderer.py”, line 67, in forward*
** perspective_correct=raster_settings.perspective_correct,**
** File “/home/kak/anaconda3/envs/gt/lib/python3.6/site-packages/pytorch3d/renderer/mesh/rasterize_meshes.py”, line 150, in rasterize_meshes**
** cull_backfaces,**
** File “/home/kak/anaconda3/envs/gt/lib/python3.6/site-packages/pytorch3d/renderer/mesh/rasterize_meshes.py”, line 205, in forward**
** cull_backfaces,**
TypeError: rasterize_meshes(): incompatible function arguments. The following argument types are supported:
** 1. (arg0: at::Tensor, arg1: at::Tensor, arg2: at::Tensor, arg3: int, arg4: float, arg5: int, arg6: int, arg7: int, arg8: bool, arg9: bool, arg10: bool) → Tuple[at::Tensor, at::Tensor, at::Tensor, at::Tensor]**
Invoked with: tensor([[[-2.9129e-01, 3.8580e-01, 9.9854e+00],
** [-3.1403e-01, 3.8133e-01, 9.9865e+00],**
** [-2.8892e-01, 3.7910e-01, 9.9895e+00]],**
** [[-3.2832e-01, 4.5427e-01, 1.0014e+01],**
** [-3.4195e-01, 4.3908e-01, 1.0008e+01],**
** [-3.3878e-01, 4.2805e-01, 1.0044e+01]],**
** [[-2.9508e-01, 2.3414e-02, 1.0420e+01],**
** [-2.4708e-01, -2.5840e-04, 1.0480e+01],**
** [-2.3614e-01, 6.3354e-02, 1.0425e+01]],**
** …,**
** [[ 8.7310e-02, 2.2173e-01, 9.9789e+00],**
** [ 9.3325e-02, 2.2365e-01, 9.9810e+00],**
** [ 9.5288e-02, 2.1816e-01, 9.9787e+00]],**
[[ 8.8948e-02, 2.1716e-01, 9.9770e+00],
[ 9.5288e-02, 2.1816e-01, 9.9787e+00],
[ 9.6179e-02, 2.1224e-01, 9.9768e+00]],
[[ 9.6179e-02, 2.1224e-01, 9.9768e+00],
[ 9.5954e-02, 2.0613e-01, 9.9754e+00],
[ 8.9500e-02, 2.0716e-01, 9.9743e+00]]], device='cuda:0',
grad_fn=<IndexBackward>), tensor([ 0, 9976, 19952, 29928, 39904, 49880, 59856, 69832, 79808,
89784, 99760, 109736, 119712, 129688, 139664, 149640, 159616, 169592,
179568, 189544], device='cuda:0'), tensor([9976, 9976, 9976, 9976, 9976, 9976, 9976, 9976, 9976, 9976, 9976, 9976,
9976, 9976, 9976, 9976, 9976, 9976, 9976, 9976], device='cuda:0'), 224.0, 0.0, 1, 16, 10000, False, False, False