This final command will build and install the package within your current environment, ready to go. On a CMD terminal at the PythonAPI folder, run: python setup.py build_ext -inplace.The rationale here is to remove the Clang specific arguments, which don’t work on MVCC. Edit line 12 to be extra_compile_args=.Navigate to the PythonAPI folder and open the setup.py file.If you already have this package installed, you can safely skip this part.
Many of TF’s Object Detection API modules depend on the pycocotools package, which fails to install on Windows via pip. At the end of this article, I provide a code snippet to run object detection on a test image, which you can later alter for your needs. In case you are not a TensorFlow user or a Data Scientist and just want to get a object detection algorithms to work, all dependencies besides Python will be installed along the way, so you don’t need to install TensorFlow yourself. The former is likely to be updated during the process, so keep this in mind if you need a specific version.īesides your Python environment, you will need a text editor and a Git client. I conducted the installation process using TensorFlow 2.3 and Python 3.7.6. I assume you are familiar with Python, is using a virtual environment of some sort, and is familiar with TensorFlow. Here, I present to you how to do it.ĭiscla imer: I am not against Docker in any way. Long story short, I rolled up my sleeves and went through it, fixing one step at a time until it worked. Unfortunately, however, the installation process is far more troublesome than it looks, and the official docs are Docker-based, which just seems an unnecessary burden for what ought to be only an ordinary pip package. Since I was using TensorFlow 2, the Object Detection API seemed a good fit. Recently, I needed to run an object detection model for a proof-of-concept I was working on. Object Detection using the code snipped provided in this tutorial
0 kommentar(er)
