AHMET SOYYIGIT

Reproducing results in the paper

4

Hello, How can we reproduce the results in the TorchSparse paper at Figure 11? Do you provide the implementation of MinkUNet and CenterPoint that utilizes TorchSparse? Also, when I look...

In file backend/hash/hash_cuda.cu, the following variable appears unused for the hash calculation: `extern __shared__ int kernel_offset_local[]; ` Am I missing something or is this variable supposed to be used in...

Hello, Is it possible to efficiently calculate the output voxel coordinates of a series of sparse convolution operations without actually running the convolution? It seems like calculating the indices pairs...

Deterministic execution time

3

I am trying to test how the convolution time changes based on input size. I see that the processing time can increase when a part of the input is filtered/reduced,...

Hello, thanks for the repository. Whenever I execute the extract method of cuCluster, it prints the following: LINE:178 18696 The number after 178 changes across calls. This happens with the...

Hi, I looked into what layers are there inside the rpn.onnx network. I can see the backbone part mentioned in the paper. However, it looks like the detection head (SSD)...

Please let me know in which part of the code the dynamic deadlines are determined by the scheduler.

Sensor data in closed-loop simulation

3

*I see that there is sensor data available to download but I am wondering what happens if we run closed-loop simulation. Does it generate simulated sensor data? I need the...

FPS drops to ~5 when Autoware starts

1

Hello, thanks for this great project. I run Autoware on Jetson AGX Orin and AWSIM on a Ubuntu 24.04 PC equipped with i7-11700KF + RTX 3080 + 64 GB DRAM....

Hello, thanks for this wonderful project. I want to train DSVT on nuscenes with different voxel sizes to have models with different execution times and accuracy tradeoffs. For example, Centerpoint...