YoloSeg memory leak

[md44]

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU) GPU
• DeepStream Version 6.3
• JetPack Version (valid for Jetson only)
• TensorRT Version
• NVIDIA GPU Driver Version (valid for GPU only)
• Issue Type( questions, new requirements, bugs)
• How to reproduce the issue ? (This is for bugs. Including which sample app is using, the configuration files content, the command line used and other details for reproducing)
• Requirement details( This is for new requirement. Including the module name-for which plugin or for which sample application, the function description)

Hi,
I’m using YoloSeg from GitHub - marcoslucianops/DeepStream-Yolo-Seg: NVIDIA DeepStream SDK 6.3 / 6.2 / 6.1.1 / 6.1 / 6.0.1 / 6.0 implementation for YOLO-Segmentation models
I’m experiencing a serious memory leak, after disabling my code I see that the memory leak persist, looking into the code from the above link I see in this file: DeepStream-Yolo-Seg/nvdsinfer_custom_impl_Yolo_seg/nvdsparseseg_Yolo.cpp at master · marcoslucianops/DeepStream-Yolo-Seg · GitHub in line 42 a heap memory allocation for the mask, I was wondering if this memory allocation causes my memory leak? if so what and where is the proper way to deallocate this buffer?

thanks.

[guyda]

Hi @md44
I need to mention this is not an official NVIDIA repo so our help her is very narrow.
Any case, it seems like development of this repo has stopped.
I found a similar issue posted there with no answer:
memory leaky · Issue #14 · marcoslucianops/DeepStream-Yolo-Seg (github.com)

You can consider writing to the repo owner or the developer who opened the issue.

Since the repo is not maintained, it is very possible that any of the following occurred:

1. I don’t know what version of YOLO are you using, but it’s possible there is a mismatch in the used version with intended version in this repo.
2. The supplied code might have never worked

My suggestion:

1. Look for other similar implementations
2. Try to implement the custom parser in CUDA, using existing parsers in the DeepStream libraries as reference.