oatml / oatomobile Goto Github PK

Hi authors,

Thank you very much for open-sourcing your code for the greater community. I am wondering if there are plans to also release codebase used for the Nuscenes portion of the paper? In particular, I am wondering if there are implementations of the models that would work with Nuscenes. As far as I know, the current repository only contains models interfacing CARLA.

1

this looks cool, and i want to use it, but there's no list of environment names, and it appears to use a custom function to make the environments ... would it be possible to use gym.make for this? we make hundreds of environments in a loop. it's ideal if they all work the same way:

env = gym.make(name)

if this project must use a custom way to make the environments, then where is the list of possible environments we can use? then we can do, env = make_carsuite(name) if name in CARSUITE_NAMES else gym.make(name)

Hello,

This work is very exciting, and I am interested in reproducing the results from the paper. However, I could not find pretrained models nor the training dataset, which would be useful for reproducing the results. Could you please provide pretrained models and/or the dataset used? Thanks!

Hey! Thank you for the amazing work. You had also talked about an adaptive robust imitative planning algorithm in the paper. Is there an implementation of this algorithm?

Hi developers,

Could you provide a simple tutorial on how to run certain benchmark?

Here's what I've tried but didn't work(folder is created but no data is being recorded, simulator also not popping up)

import functools
from oatomobile.benchmarks.carnovel.benchmark import carnovel
from oatomobile.baselines.rulebased.autopilot.agent import AutopilotAgent

agent_fn = functools.partial(
    AutopilotAgent,
    # More configs
)

carnovel.evaluate(agent_fn, log_dir='log/', subtasks_id='AbnormalTurns0-v0', monitor=False, render=False)

Hello.

I'm trying to visualize the predictions of DIM agent. I was able to do so by using the carnovel benchmark. But in the output videos, the car does not exactly follow the prediction.

Here's a sample of the generated outputs. As you can see in the video, the car is not following the red line.
dim.zip

Is there a constraint that prevents the car from getting out of lane?

Hi, great repo! Wondering of it's compatible with the latest Carla? Getting some segfaults:

pygame 1.9.6
Hello from the pygame community. https://www.pygame.org/contribute.html
4.24.3-0+++UE4+Release-4.24 518 0
Disabling core dumps.
Signal 11 caught.
Malloc Size=65538 LargeMemoryPoolOffset=65554
CommonUnixCrashHandler: Signal=11
Malloc Size=65535 LargeMemoryPoolOffset=131119
Malloc Size=118144 LargeMemoryPoolOffset=249280
Engine crash handling finished; re-raising signal 11 for the default handler. Good bye.
Signal 11 caught.
Malloc Size=65538 LargeMemoryPoolOffset=65554
CommonUnixCrashHandler: Signal=11
Malloc Size=65535 LargeMemoryPoolOffset=131119
Malloc Size=105648 LargeMemoryPoolOffset=236784
Engine crash handling finished; re-raising signal 11 for the default handler. Good bye.
Segmentation fault (core dumped)


4.24.3-0+++UE4+Release-4.24 518 0
Disabling core dumps.
Signal 11 caught.
Malloc Size=65538 LargeMemoryPoolOffset=65554
CommonUnixCrashHandler: Signal=11
Malloc Size=65535 LargeMemoryPoolOffset=131119
Malloc Size=118144 LargeMemoryPoolOffset=249280
Engine crash handling finished; re-raising signal 11 for the default handler. Good bye.
Signal 11 caught.
Malloc Size=65538 LargeMemoryPoolOffset=65554
CommonUnixCrashHandler: Signal=11
Malloc Size=65535 LargeMemoryPoolOffset=131119
Malloc Size=105648 LargeMemoryPoolOffset=236784
Engine crash handling finished; re-raising signal 11 for the default handler. Good bye.
Segmentation fault (core dumped)
4.24.3-0+++UE4+Release-4.24 518 0
Disabling core dumps.
Signal 11 caught.
Malloc Size=65538 LargeMemoryPoolOffset=65554
CommonUnixCrashHandler: Signal=11
Malloc Size=65535 LargeMemoryPoolOffset=131119
Malloc Size=118144 LargeMemoryPoolOffset=249280
Engine crash handling finished; re-raising signal 11 for the default handler. Good bye.
Signal 11 caught.
Malloc Size=65538 LargeMemoryPoolOffset=65554
CommonUnixCrashHandler: Signal=11
Malloc Size=65535 LargeMemoryPoolOffset=131119
Malloc Size=105648 LargeMemoryPoolOffset=236784
Engine crash handling finished; re-raising signal 11 for the default handler. Good bye.
Segmentation fault (core dumped)

Thanks!

I can't find any code setting weather conditions.
Is there any way to collect demonstrations with specific weather parameters?

Hello, I am very interested in your work, and this code is very helpful to me. However, I still have some issues about oatomobile/oatomobile/torch/networks/sequence.py

1. The default output_shape of AutoregressiveFlow is (4,2), and the trajectory consisting of 40 points is obtained by interpolating the 4 output points. Why not just output more points？
2. If the output_shape of AutoregressiveFlow is not (4,2), e.g., (16,2), the output_sizes of self._locscale may be not right, and the following code will error out:

dloc_scale = self._locscale(z)
dloc = dloc_scale[..., :2]
scale = F.softplus(dloc_scale[..., 2:]) + 1e-3
y_t = (y_tm1 + dloc) + scale * x_t

So I change the output_sizes of self._locscale to a fixed [32, 4], but I don't know if this is appropriate.
3. I'm not very familiar with Autoregressive Flow, so I'm confused about the process of calculating y and logabsdet. I would appreciate it if you could tell me which paper has a description of this part of the calculation process.

Hi! Thanks for your code!
I met a problem when I run 'pip install oatomobile[torch]' it reported

Collecting oatomobile[torch]
Using cached oatomobile-0.1.0-py3-none-any.whl
ERROR: Cannot install oatomobile and oatomobile[torch]==0.1.0 because these package versions have conflicting dependencies.

The conflict is caused by:
oatomobile[torch] 0.1.0 depends on torch==1.5.1; extra == "torch"
torchvision 0.6.0 depends on torch==1.5.0

To fix this you could try to:

1. loosen the range of package versions you've specified
2. remove package versions to allow pip attempt to solve the dependency conflict

ERROR: ResolutionImpossible: for help visit https://pip.pypa.io/en/latest/user_guide/#fixing-conflicting-dependencies

I am trying to collect dataset using baselines/rulebased/autopilot/run.py. It worked but after I added "player_future" in the sensors list it still did not produce the player_future files. Looks like there is no player_future implementation?

I'm getting an error when running "pip install oatomobile".

Here is my command prompt output:

C:\Users\david>pip install oatomobile --log oatlog.txt
Collecting oatomobile
  Using cached oatomobile-0.1.0.tar.gz (62 kB)
Collecting absl-py==0.9.0
  Using cached absl-py-0.9.0.tar.gz (104 kB)
Collecting gym==0.17.1
  Using cached gym-0.17.1.tar.gz (1.6 MB)
Collecting pygame==1.9.6
  Using cached pygame-1.9.6.tar.gz (3.2 MB)
ERROR: Command errored out with exit status 1: python setup.py egg_info Check the logs for full command output.

C:\Users\david>pip list
Package    Version
---------- -------
numpy      1.19.4
pip        20.2.4
pygame     2.0.0
setuptools 50.3.2

And attached is the logfile generated by the installation attempt.

The key error might be "ERROR: Command errored out with exit status 1: python setup.py egg_info Check the logs for full command output." But I'm not experienced with pip and don't know what this really means.

oatlog.txt

Hi,
Is there any way to collect camera images instead of LiDAR, and train perception module with them?