haomo-ai / cam4docc Goto Github PK
View Code? Open in Web Editor NEW[CVPR 2024] Cam4DOcc: Benchmark for Camera-Only 4D Occupancy Forecasting in Autonomous Driving Applications
License: MIT License
[CVPR 2024] Cam4DOcc: Benchmark for Camera-Only 4D Occupancy Forecasting in Autonomous Driving Applications
License: MIT License
We sincerely look forward to your poster and will arrange a collection (will show on bilibili)for all authors as a pre-heat for the offline conference. It is convenient for everyone to discuss the posters in advance and this will be a period of significant memory for everyone. You can also send us the exhibition information with poster.
Our email :[email protected]
WeChat:AIDriver002
I experience the following problem when using config OCFNet_in_Cam4DOcc_V1.2.py
:
File "/home/eitan/Cam4DOcc/projects/occ_plugin/occupancy/image2bev/ViewTransformerLSSBEVDepth.py", line 495, in forward
depth = self.depth_conv(depth)
File "/fs/scratch/rb_bd_dlp_rng-dl01_cr_AID_employees/users/eitan/envs/cam4docc/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl
return forward_call(*input, **kwargs)
File "/fs/scratch/rb_bd_dlp_rng-dl01_cr_AID_employees/users/eitan/envs/cam4docc/lib/python3.7/site-packages/torch/nn/modules/container.py", line 141, in forward
input = module(input)
File "/fs/scratch/rb_bd_dlp_rng-dl01_cr_AID_employees/users/eitan/envs/cam4docc/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1102, in _call_impl
return forward_call(*input, **kwargs)
File "/fs/scratch/rb_bd_dlp_rng-dl01_cr_AID_employees/users/eitan/envs/cam4docc/lib/python3.7/site-packages/mmcv/ops/deform_conv.py", line 378, in forward
False, self.im2col_step)
File "/fs/scratch/rb_bd_dlp_rng-dl01_cr_AID_employees/users/eitan/envs/cam4docc/lib/python3.7/site-packages/mmcv/ops/deform_conv.py", line 109, in forward
im2col_step=cur_im2col_step)
RuntimeError: CUDA error: no kernel image is available for execution on the device
What could be the problem? The installation seemed to work fine without any error.
Hi! I noticed that your script uses 8 GPUs for training and testing simultaneously. How many GPU resources are needed to reproduce the results of this task, training and testing, in totally? Thanks for noticing this issue!
Hi, thanks to your great work,
Can you tell me how much GPU memory is needed approximately for this task?
In line 293 of ocfnet.py, I think the transformed_grid.shape[1]=4, which may indicate you only wrap 4 voxel grid feature to the wraped_x from x. I don't know if I understand right, I think the for loop should be the number of voxels. May you check again this part?
Traceback (most recent call last):
File "tools/train.py", line 204, in <module>
main()
File "tools/train.py", line 200, in main
meta=meta)
File "/home/glj/roseupram/Cam4DOcc/projects/occ_plugin/occupancy/apis/train.py", line 27, in custom_train_model
meta=meta)
File "/home/glj/roseupram/Cam4DOcc/projects/occ_plugin/occupancy/apis/mmdet_train.py", line 113, in custom_train_detector
runner.run(data_loaders, cfg.workflow)
File "/home/glj/miniconda3/envs/cam4docc/lib/python3.7/site-packages/mmcv/runner/epoch_based_runner.py", line 127, in run
epoch_runner(data_loaders[i], **kwargs)
File "/home/glj/miniconda3/envs/cam4docc/lib/python3.7/site-packages/mmcv/runner/epoch_based_runner.py", line 47, in train
for i, data_batch in enumerate(self.data_loader):
File "/home/glj/miniconda3/envs/cam4docc/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 521, in __next__
data = self._next_data()
File "/home/glj/miniconda3/envs/cam4docc/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 1203, in _next_data
return self._process_data(data)
File "/home/glj/miniconda3/envs/cam4docc/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 1229, in _process_data
data.reraise()
File "/home/glj/miniconda3/envs/cam4docc/lib/python3.7/site-packages/torch/_utils.py", line 434, in reraise
raise exception
FileNotFoundError: Caught FileNotFoundError in DataLoader worker process 0.
Original Traceback (most recent call last):
File "/home/glj/miniconda3/envs/cam4docc/lib/python3.7/site-packages/torch/utils/data/_utils/worker.py", line 287, in _worker_loop
data = fetcher.fetch(index)
File "/home/glj/miniconda3/envs/cam4docc/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 49, in fetch
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/home/glj/miniconda3/envs/cam4docc/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 49, in <listcomp>
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/home/glj/roseupram/Cam4DOcc/projects/occ_plugin/datasets/cam4docc_dataset.py", line 100, in __getitem__
data = self.prepare_train_data(idx)
File "/home/glj/roseupram/Cam4DOcc/projects/occ_plugin/datasets/cam4docc_dataset.py", line 312, in prepare_train_data
example = self.prepare_sequential_data(index)
File "/home/glj/roseupram/Cam4DOcc/projects/occ_plugin/datasets/cam4docc_dataset.py", line 370, in prepare_sequential_data
example = self.pipeline(input_seq_data)
File "/home/glj/miniconda3/envs/cam4docc/lib/python3.7/site-packages/mmdet/datasets/pipelines/compose.py", line 40, in __call__
data = t(data)
File "/home/glj/roseupram/Cam4DOcc/projects/occ_plugin/datasets/pipelines/loading_instance.py", line 392, in __call__
np.savez(seg_label_path, segmentation_saved_list2)
File "<__array_function__ internals>", line 6, in savez
File "/home/glj/miniconda3/envs/cam4docc/lib/python3.7/site-packages/numpy/lib/npyio.py", line 616, in savez
_savez(file, args, kwds, False)
File "/home/glj/miniconda3/envs/cam4docc/lib/python3.7/site-packages/numpy/lib/npyio.py", line 712, in _savez
zipf = zipfile_factory(file, mode="w", compression=compression)
File "/home/glj/miniconda3/envs/cam4docc/lib/python3.7/site-packages/numpy/lib/npyio.py", line 112, in zipfile_factory
return zipfile.ZipFile(file, *args, **kwargs)
File "/home/glj/miniconda3/envs/cam4docc/lib/python3.7/zipfile.py", line 1240, in __init__
self.fp = io.open(file, filemode)
FileNotFoundError: [Errno 2] No such file or directory: './data/segmentation/2175d0e84f224ea69907e5c338bde395_b60a2bc82d88419497510082d9301a59.npz'
where is segmentation folder?
The amount of data in v1.0-trainval is too large
Hello! I have read this paper, which is an impressive work. However, I want to run your code on my computer (RTX4090) because the default grid size is 512 x 512 x 40, which is too big.
I have tried to change the grid_size to 200 x 200 x 16 and it can run appropriately. However, when I run the vis_gt.py, I get the below result:
thanks for your release
I only have one gpu, memory is 48G, so i run this
run.sh ./projects/configs/baselines/OCFNet_in_Cam4DOcc_V1.1.py 1
but seems my gpu is not enough
RuntimeError: CUDA out of memory. Tried to allocate 198.00 MiB (GPU 0; 47.54 GiB total capacity; 45.25 GiB already allocated; 18.19 MiB free; 45.46 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF
Is there a way to run this code with one 48G gpu
Hi,How long does it take to generate the dataset?
Hi, could you provide the 4D annotations as zip files?
Generating the dataset is taking longer than a week.
Thanks
Dear Authors,
I noticed that the numerical results provided in your paper seem to have been obtained using depth loss. However, the source code you have provided does not include the implementation of depth loss. Could you please clarify if the results in the paper were indeed obtained using depth loss, and if so, could you provide the implementation details or update the source code to include it?
Thank you!
Best regards,
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.