hamihamiha
hamihamiha
If you don't want to retrain a model and only want to detect one class or several classes. Maybe you can try this method #204 .
I also want to know if there are some methods to detect only one class or several classes? #158 says that we should to retrain a model, but I think...
I found a way to solve this question. In the YoloObjectDetector.cpp, I found this function `draw_detections(display, dets, nboxes, demoThresh_, demoNames_, demoAlphabet_, demoClasses_);` This function is used to draw BB and...
Sorry, cause video may have differt FPS. I hope it not confuse you, and you can use multi process to accelerate the process speed.
Hi, could you tell me which file you changed? I have the same question.
您好，可以看一下segmentation_models_pytorch的版本吗
Hello, @BrunoGeorgevich Could you show your initialize code, I use this model and it will increase memory continously.
Hi @BrunoGeorgevich Thank you, Your code has inspired me and I found my problem. I save the everything_results in a dictionary which kept increasing the GPU memory. I have placed...
Hello @karaposu ， I think it's not the image size affects the inference time, but loading the model to gpu will consume more time. Can you change the image order...
还有一个问题，在进行视频风格转换的过程中，会出现眼部抖动的问题，我怀疑是以下问题导致的，请问有没有什么解决思路？ ```python cannot find 5 points with find_puil, used mean instead.! ``` 我现在想的是跳过这些帧，但是modelscope的代码封装的太好了，我找不到源代码，不知道如何修改