augmentedstartups / as-one Goto Github PK

Dear All,

I am facing installation issues. More specific when i install numpy version 1.24 i am no facing issues with Lapjv but i am receiving error with floats. if i downgrade the numpy to 1.22 version i am facing the subject code error. with version 1.24 numpy the windows open recognize the first car but after fails with subject error. Your support will be appreciated.

Kind regards

Hello,

I am trying to install this software on Linux, however, I get weird error:

yolov5_detector.py", line 8, in
import coremltools as ct
ModuleNotFoundError: No module named 'coremltools'

I am under the impression that coreml are for Apple machines.

Hello,

When using ByteTrack tracker I encountered and issue that the tracked objects and class ids / class names are sometimes mismatched. I see that in the detect_and_track function, all class ids are extracted and all detections are sent to the _tracker_update, but the tracker update filters out some of the targets with the box shape criteria resulting in a difference between numbers of bboxes_xyxy, ids and scores and number of class_ids returned by the detect_and_track function. Could this cause the error while displaying detections since the index of each bounding box does not match the index of the class id and ultimately class name?
If so, could a solution be to modify the _tracker_update to add an online_class_ids list and append the class ids extracted from dets, for each online target that passes the box shape criteria and return that list along with online_xyxys, online_ids, online_scores and use it later instead of the original class_ids?

Thanks in advance!

Integrate grounding dino and segment anything with AS-One
https://github.com/IDEA-Research/Grounded-Segment-Anything

Hi everyone,
I found this repository, and it works perfectly with downloaded videos and fotos, but I have a question. Do you know how to use this repository with my pc camera or an external camera?

Thanks in advance

hi,
I'm running the detector on a custom model (i.e., i provide the the weights and classes)
got an error in line 107
warnings.warn(
f"class {_class} not found in model classes list.")

I can resolve it when I replace line 99
class_names = get_names()
with
class_names = list(self.model.names.values())

Am I right?
thanks

Hi, Can you please let me know how can we detect and track only humans in a video sequence?

Hi,

Just wondering if I have to build my entire computer vision inside the AS-One project, or is there any simple way of exporting the detector I've created?

Thanks

In "detect" method you have created "detections" variable and instead of it you use "detection".

def detect(self, 
               image: list,
               conf_thres: float = 0.25,
               iou_thres: float = 0.45,
               classes: int = None,
               agnostic_nms: bool = False,
               input_shape=(640, 640),
               max_det: int = 1000,
               filter_classes = None) -> list:
     
        # Image Preprocessing
        original_image, processed_image = self.image_preprocessing(image, input_shape)
        
        # Inference
        if self.use_onnx:
            # Input names of ONNX model on which it is exported   
            input_name = self.model.get_inputs()[0].name
            # Run onnx model 
            pred = self.model.run([self.model.get_outputs()[0].name], {input_name: processed_image})[0]
            # Run Pytorch model        
        else:
            processed_image = torch.from_numpy(processed_image).to(self.device)
            # Change image floating point precision if fp16 set to true
            processed_image = processed_image.half() if self.fp16 else processed_image.float() 
            pred = self.model(processed_image, augment=False, visualize=False)[0]
       
        # Post Processing
        if isinstance(pred, np.ndarray):
            pred = torch.tensor(pred, device=self.device)
        predictions = non_max_suppression(pred, conf_thres, 
                                          iou_thres, classes, 
                                          agnostic_nms, 
                                          max_det=max_det)
        
        for i, prediction in enumerate(predictions):  # per image
            if len(prediction):
                prediction[:, :4] = scale_coords(
                    processed_image.shape[2:], prediction[:, :4], original_image.shape).round()
                predictions[i] = prediction
        detections = predictions[0].cpu().numpy()
        image_info = {
            'width': original_image.shape[1],
            'height': original_image.shape[0],
        }

        self.boxes = detections[:, :4]
        self.scores = detections[:, 4:5]
        self.class_ids = detections[:, 5:6]

        if filter_classes:
            class_names = get_names()

            filter_class_idx = []
            if filter_classes:
                for _class in filter_classes:
                    if _class.lower() in class_names:
                        filter_class_idx.append(class_names.index(_class.lower()))
                    else:
                        warnings.warn(f"class {_class} not found in model classes list.")

            detection = detection[np.in1d(detection[:,5].astype(int), filter_class_idx)]

        return detections, image_info

File "/home/zhora/workspace/AS-One/asone/detectors/yolov5/yolov5_detector.py", line 118, in detect
    detection = detection[np.in1d(detection[:,5].astype(int), filter_class_idx)]
UnboundLocalError: local variable 'detection' referenced before assignment

Support for strong sort will be very helpful, options can be given like whether to use strong sort or deep sort.

Hi thanks for great repository. I want to use only detection, not tracking. And also want to see class names. Help please

I train a model for pigs and other animals. (yolov5x)
so how to use my weight pt file to this model?
i want to track and counting function.

thanks

Hello,
I try to build a docker container for asone based app, but I get some strange behavior.
First, the error:
root@9c83bf0cbc7b:/usr/src/app# python3
Python 3.8.10 (default, Nov 14 2022, 12:59:47)
[GCC 9.4.0] on linux
Type "help", "copyright", "credits" or "license" for more information.

import asone
Traceback (most recent call last):
File "", line 1, in
File "/usr/local/lib/python3.8/dist-packages/asone/init.py", line 1, in
from .asone import ASOne
File "/usr/local/lib/python3.8/dist-packages/asone/asone.py", line 6, in
import asone.utils as utils
File "/usr/local/lib/python3.8/dist-packages/asone/utils/init.py", line 6, in
from asone.utils.temp_loader import get_detector, get_tracker
File "/usr/local/lib/python3.8/dist-packages/asone/utils/temp_loader.py", line 1, in
from asone.detectors import YOLOv5Detector
File "/usr/local/lib/python3.8/dist-packages/asone/detectors/init.py", line 7, in
from asone.detectors.detector import Detector
File "/usr/local/lib/python3.8/dist-packages/asone/detectors/detector.py", line 13, in
from asone.detectors.easyocr_detector.text_detector import TextDetector
ModuleNotFoundError: No module named 'asone.detectors.easyocr_detector'

second, ls on asone distro show no easyocr_detector folder. (The folder exist on the github.com/augmentedstartups/AS-One)
root@9c83bf0cbc7b:/usr/src/app# ls /usr/local/lib/python3.8/dist-packages/asone/detectors
init.py pycache detector.py utils yolor yolov5 yolov6 yolov7 yolov8 yolox

I used "pip install asone".
I guess it is a misconfigured infos between PyPI and github!?

Hello,

Is there a simple way to install this software on Linux? I face several incompatibility issue that are hard to resolve. Which version of super_gradient, numpy this software requires?

how can i fix this error please !

Hello,
Thank you for this amazing work!
I have some installation issues, I have this error while installing asone, or even doing it separately (for example here for pip install lap)

`INFO:

  ########### EXT COMPILER OPTIMIZATION ###########
  INFO: Platform      :
    Architecture: x64
    Compiler    : msvc

  CPU baseline  :
    Requested   : 'min'
    Enabled     : none
    Flags       : none
    Extra checks: none

  CPU dispatch  :
    Requested   : 'max -xop -fma4'
    Enabled     : none
    Generated   : none
  INFO: CCompilerOpt.cache_flush[825] : write cache to path -> C:\Users\AppData\Local\Temp\pip-install-etdq5zeh\lap_45220a53799847869a905004fb7a2314\build\temp.win-amd64-cpython-38\Release\ccompiler_opt_cache_ext.py

note: This error originates from a subprocess, and is likely not a problem with pip.
ERROR: Failed building wheel for lap
Running setup.py clean for lap
Failed to build lap
Installing collected packages: lap
Running setup.py install for lap ... error
error: subprocess-exited-with-error

× Running setup.py install for lap did not run successfully.
│ exit code: 1
╰─> [107 lines of output]
Partial import of lap during the build process.
Generating cython files
running install
C:\Users\anaconda3\envs\py38_GPU_pyqt\lib\site-packages\setuptools\command\install.py:34: SetuptoolsDeprecationWarning: setup.py install is deprecated. Use build and pip and other standards-based tools.
warnings.warn(
running build
running config_cc
INFO: unifing config_cc, config, build_clib, build_ext, build commands --compiler options
running config_fc
INFO: unifing config_fc, config, build_clib, build_ext, build commands --fcompiler options
running build_src
INFO: build_src
INFO: building extension "lap.lapjv" sources
INFO: building data_files sources
INFO: build_src: building npy-pkg config files
running build_py
creating build
creating build\lib.win-amd64-cpython-38
creating build\lib.win-amd64-cpython-38\lap
copying lap\lapmod.py -> build\lib.win-amd64-cpython-38\lap
copying lap_init.py -> build\lib.win-amd64-cpython-38\lap
running build_ext
INFO: No module named 'numpy.distutils._msvccompiler' in numpy.distutils; trying from distutils
INFO: customize MSVCCompiler
INFO: customize MSVCCompiler using build_ext
INFO: CCompilerOpt.cc_test_flags[1029] : testing flags (/O2)
creating C:\Users\CHOUCH1\AppData\Local\Temp\tmphagtbu1r\Users
creating C:\Users\CHOUCH1\AppData\Local\Temp\tmphagtbu1r\Users\chouchen2-admin
creating C:\Users\CHOUCH1\AppData\Local\Temp\tmphagtbu1r\Users\chouchen2-admin\anaconda3
creating C:\Users\CHOUCH1\AppData\Local\Temp\tmphagtbu1r\Users\chouchen2-admin\anaconda3\envs
creating C:\Users\CHOUCH1\AppData\Local\Temp\tmphagtbu1r\Users\chouchen2-admin\anaconda3\envs\py38_GPU_pyqt
creating C:\Users\CHOUCH1\AppData\Local\Temp\tmphagtbu1r\Users\chouchen2-admin\anaconda3\envs\py38_GPU_pyqt\Lib
creating C:\Users\CHOUCH1\AppData\Local\Temp\tmphagtbu1r\Users\chouchen2-admin\anaconda3\envs\py38_GPU_pyqt\Lib\site-packages
creating C:\Users\CHOUCH1\AppData\Local\Temp\tmphagtbu1r\Users\chouchen2-admin\anaconda3\envs\py38_GPU_pyqt\Lib\site-packages\numpy
creating C:\Users\CHOUCH1\AppData\Local\Temp\tmphagtbu1r\Users\chouchen2-admin\anaconda3\envs\py38_GPU_pyqt\Lib\site-packages\numpy\distutils
creating C:\Users\CHOUCH1\AppData\Local\Temp\tmphagtbu1r\User\anaconda3\envs\py38_GPU_pyqt\Lib\site-packages\numpy\distutils\checks
INFO: CCompilerOpt.cc_test_flags[1029] : testing flags (/WX)
WARN: CCompilerOpt.init[1175] : feature 'AVX512_KNL' is disabled, MSVC compiler doesn't support it
WARN: CCompilerOpt.init[1175] : feature 'AVX512_KNM' is disabled, MSVC compiler doesn't support it
INFO: CCompilerOpt.init[1717] : check requested baseline
INFO: CCompilerOpt.feature_test[1482] : testing feature 'SSE' with flags ()
WARN: CCompilerOpt.dist_test[598] : CCompilerOpt._dist_test_spawn[732] : Command (C:\Program Files (x86)\Microsoft Visual Studio\2019\BuildTools\VC\Tools\MSVC\14.28.29910\bin\HostX86\x64\cl.exe /c /nologo /O2 /W3 /GL /DNDEBUG /MD -IC:\Users\anaconda3\envs\py38_GPU_pyqt\include -IC:\Users\chouchen2-admin\anaconda3\envs\py38_GPU_pyqt\Include -IC:\Program Files (x86)\Microsoft Visual Studio\2019\BuildTools\VC\Tools\MSVC\14.28.29910\include -IC:\Program Files (x86)\Microsoft Visual Studio\2019\BuildTools\VC\Tools\MSVC\14.28.29910\include -IC:\Program Files (x86)\Microsoft Visual Studio\2019\BuildTools\VC\Tools\MSVC\14.28.29910\include /TcC:\Users\chouchen2-admin\anaconda3\envs\py38_GPU_pyqt\Lib\site-packages\numpy\distutils\checks\cpu_sse.c /FoC:\Users\CHOUCH~1\AppData\Local\Temp\tmphagtbu1r\Users\anaconda3\envs\py38_GPU_pyqt\Lib\site-packages\numpy\distutils\checks\cpu_sse.obj /WX) failed with exit status 2 output

`

So, could you tell me if it is related to the Virtual Studio version? Thanks

how can i print center a object. Please i am a student !!!

I get this error when trying to run it in colab

regards
Nico

I have been trying to follow the steps provided to get this working but not able to, could you provide updated guidance on how to get this project running.

when i run the code on linux/windows/google colab, i meet same mistakes:

Hello guys, Thank you for such amazing work!

Is there a way to specify a custom path for downloaded weights instead of the current working dir ?
e.g: to be able to use the Tracker this way

Tracker(asone.BYTETRACK, model, use_cuda=False, custom_model_path)

in case the custom_model_path already exists the .pts are loaded from that path (the same way it's downloaded to the working dir at the moment).

Thanks

When going through the installation instructions and running $pip install typing_extensions==4.7.1, I get:

Collecting typing_extensions==4.7.1
Downloading typing_extensions-4.7.1-py3-none-any.whl.metadata (3.1 kB)
Downloading typing_extensions-4.7.1-py3-none-any.whl (33 kB)
Installing collected packages: typing_extensions
Attempting uninstall: typing_extensions
Found existing installation: typing-extensions 3.10.0.2
Uninstalling typing-extensions-3.10.0.2:
Successfully uninstalled typing-extensions-3.10.0.2

ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
asone 0.3.3 requires typing-extensions==3.10.0.2, but you have typing-extensions 4.7.1 which is incompatible.

Successfully installed typing_extensions-4.7.1

When I attempt to 'pip install asone', I get error with the lap package
Has anyone else seen this issue? Is there a fix?

this appear when running the collab notebook

Are you going to add OCSORT (https://github.com/noahcao/OC_SORT) as a tracker module in the future? Considering its speed it would be interesting!

Hi,

Will the trackers: Deep OC-SORT and SMILEtrack be added?

Thanks!

Hi,
First thank you for a great git, I have a question (This is not an issue) this project can run on Jetson Xavier.
Did you try this project on some Nvidia Jetson?
Do I need to install something special to run this project on Nvidia Jetson or it's the same libraries?

Thank you again :)

Hello,

it seems there are some issues in installating cython-bbox on Apple M1. Is there any alternative or solution to the compilation
ussue so that AsOne can be installed?

Thansk

Hello. Would it be possible to integrate TensorRT support in the library for at least the models that support it? I have a yolov7 model to run on a Jetson Nano, and i would like the faster inference offered by tensorrt, while getting the accesibility and simplicity that this wonderful module brings. Thanks!

Hello,
Thanks a lot for this amazing implementation. My question is can we easily integrate yolo-world in As-One?
Do you have any future plan to do that soon ?

Hi,

Will this software work on Mac as well?

can't open google colab

I tried to install but I encountered an error. I downloaded the MS Build Tools, but I do not know which package to install. (The interface is in Turkish)

https://github.com/WongKinYiu/yolov9

Is it possible to get a txt file with the coordinates of the detected objects? Maybe in the classical form: , , <bb_left>, <bb_top>, <bb_width>, <bb_height>, , , ,

On the yolo packages (v5,v6) we can open multiple streams by doing below:
python3 track.py --source streams.txt

And this would open multiple streams, where streams.txt looks like below:
video1.mp4
video2.mp4
video3.mp4

But AS-ONE main.py only can open one stream at a time? will multiple streams be supported in the near future?

Traceback (most recent call last):
File "main.py", line 40, in
main(args)
File "main.py", line 6, in main
dt_obj = ASOne(
File "/home/zhora/workspace/AS-One/asone/asone.py", line 20, in init
self.detector = self.get_detector(detector)
File "/home/zhora/workspace/AS-One/asone/asone.py", line 25, in get_detector
detector = Detector(detector, use_cuda=self.use_cuda).get_detector()
File "/home/zhora/workspace/AS-One/asone/detectors/detector.py", line 17, in init
self.model = self._select_detector(model_flag, use_cuda)
File "/home/zhora/workspace/AS-One/asone/detectors/detector.py", line 25, in _select_detector
_detector = YOLOv5Detector(weights=weight,
File "/home/zhora/workspace/AS-One/asone/detectors/yolov5/yolov5_detector.py", line 26, in init
self.model = self.load_model(use_cuda, weights)
File "/home/zhora/workspace/AS-One/asone/detectors/yolov5/yolov5_detector.py", line 40, in load_model
model = attempt_load(weights, device=self.device, inplace=True, fuse=True)
File "/home/zhora/workspace/AS-One/asone/detectors/yolov5/yolov5/models/experimental.py", line 33, in attempt_load
ckpt = torch.load(w, map_location='cpu') # load
File "/home/zhora/workspace/venv/lib/python3.8/site-packages/torch/serialization.py", line 712, in load
return _load(opened_zipfile, map_location, pickle_module, **pickle_load_args)
File "/home/zhora/workspace/venv/lib/python3.8/site-packages/torch/serialization.py", line 1049, in _load
result = unpickler.load()
File "/home/zhora/workspace/venv/lib/python3.8/site-packages/torch/serialization.py", line 1042, in find_class
return super().find_class(mod_name, name)
AttributeError: Can't get attribute 'DetectionModel' on <module 'models.yolo' from '/home/zhora/workspace/AS-One/asone/detectors/yolov5/yolov5/models/yolo.py'>