Jul-15-2025, 11:10 AM
I want to develop a toolkit that integrates image denoising algorithms, allowing users to select different algorithms to denoise images by themselves. Here are my questions:
However, different algorithms require different environments. How can I handle all environments within the application so that users don't need to configure the environments themselves? Are there any other methods besides using Anaconda to manage all versions and Docker containerization? How to solve the problem that the generated executable file (.exe) is too large?
I have tried packaging the environment for each algorithm separately using Conda, but the generated . The main version issue is that some algorithms use Python 2.7 and PyTorch 0.41, while others use Python 3.8 and PyTorch 1.13.1. Additionally, there are differences in the versions of OpenCV and TensorFlow.
However, different algorithms require different environments. How can I handle all environments within the application so that users don't need to configure the environments themselves? Are there any other methods besides using Anaconda to manage all versions and Docker containerization? How to solve the problem that the generated executable file (.exe) is too large?
I have tried packaging the environment for each algorithm separately using Conda, but the generated . The main version issue is that some algorithms use Python 2.7 and PyTorch 0.41, while others use Python 3.8 and PyTorch 1.13.1. Additionally, there are differences in the versions of OpenCV and TensorFlow.
