GithubHelp home page GithubHelp logo

licksylick / autotrackanything Goto Github PK

View Code? Open in Web Editor NEW
60.0 4.0 6.0 96.55 MB

AutoTrackAnything is a universal, flexible and interactive tool for insane automatic object tracking over thousands of frames. It is developed upon XMem, Yolov8 and MobileSAM (Segment Anything), can track anything which detect Yolov8.

License: MIT License

Python 95.38% C++ 1.78% Cuda 2.34% Cython 0.50%
multi-object-tracker multi-object-tracking re-id re-identification reid sam segment-anything track-anything tracking tracking-by-detection

autotrackanything's People

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar  avatar  avatar  avatar

autotrackanything's Issues

[Question] bbox to mobile sam

If I understand the project correctly the use of yolo to create bbox and then prompt sam with the bbox is similar in principle to mobilesamv2 correct?

Just trying to make sure my underlying assumptions are correct here

Pinned versions of packages cause installation problems, download script does not work

Здорова!

thank you for the project!
I am wondering why it was necessary to pin some packages to specific versions? On Fedora 38 it lead to an installation error due to incompatible dependency versions. I relaxed the version requirements like below and was then able to install and run everything just fine on my ROCm setup:

diff --git a/requirements.txt b/requirements.txt
index d9fd103..6ef930c 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -1,9 +1,9 @@
 mega.py
 mobile-sam @ git+https://github.com/ChaoningZhang/MobileSAM.git
-numpy==1.23.1
+numpy>=1.23.1
 opencv-python>=4.1.1
-pandas==1.4.3
+pandas>=1.4.3
 Pillow>=7.1.2
 scikit-image==0.19.3
-timm==0.9.12
-ultralytics
\ No newline at end of file
+timm>=0.9.12
+ultralytics

The next issue I had was with the download script, the "mega" module does not seem to work on my Python version:

Traceback (most recent call last):
  File "/mnt/ssd/xxx/AutoTrackAnything/download_models.py", line 1, in <module>
    from mega import Mega
  File "/home/xxx/.local/lib/python3.11/site-packages/mega/__init__.py", line 1, in <module>
    from .mega import Mega  # noqa
    ^^^^^^^^^^^^^^^^^^^^^^
  File "/home/xxx/.local/lib/python3.11/site-packages/mega/mega.py", line 18, in <module>
    from tenacity import retry, wait_exponential, retry_if_exception_type
  File "/home/xxx/.local/lib/python3.11/site-packages/tenacity/__init__.py", line 451, in <module>
    from tenacity._asyncio import AsyncRetrying
  File "/home/xxx/.local/lib/python3.11/site-packages/tenacity/_asyncio.py", line 33, in <module>
    class AsyncRetrying(BaseRetrying):
  File "/home/xxx/.local/lib/python3.11/site-packages/tenacity/_asyncio.py", line 41, in AsyncRetrying
    @asyncio.coroutine
     ^^^^^^^^^^^^^^^^^
AttributeError: module 'asyncio' has no attribute 'coroutine'. Did you mean: 'coroutines'?

I downloaded and unzipped the models zip manually, which worked. Still, I was wondering - why mega.nz, why not just use the one from the mobile sam repo?
https://github.com/ChaoningZhang/MobileSAM/blob/master/weights/mobile_sam.pt

I think XMem.pth should also be available from the original authors.

Other than that - really nice, it worked well on a video containing a person.

Do you have any hints on how to use it to track specific animals? Would it be enough to change the keypoint keywords or is there more to it?

How does ReID work in your project?

Hi,

Thank you for this great project! I wonder how did your project handle ReID, i.e., re-tracking the object out of frame? Can it handle re-tracking an object being out of frame for a long time?

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.