-
Hello every one, the [repo](https://github.com/linghu8812/tensorrt_inference) which support yolov4: AlexeyAB/darknet#7002, scaled yolov4: WongKinYiu/ScaledYOLOv4#56, yolov5: ultralytics/yolov5#1597, a…
-
-
大佬你好,我跑通了你的Scaled Yolov4项目。有个问题想要请教一下。
在运行
```./ScaledYOLOv4_trt ../config-p5.yaml ../samples/```
以后,生成了```yolov4-p5.trt``` engine文件。
我现在想要使用```yolov4-p5.trt```做推理,所以我写了如下代码:
```
def d…
-
Hi, Thanks for the great work!
There is one thing I would like to ask, for I would like to test the model using raw images instead of testing using dataloader;
Hence when I trying to modify the test…
-
Does the scaled yolov4 implementation in pytorch support the option of freezing layers during training?
-
NMS is taking > 1 second. Is there anyway to make it fast?
-
What annotation format should I use when training Scaled YOLOv4? COCO doesn't seem to work and shows this error:
![image](https://user-images.githubusercontent.com/87596613/133457346-a5e49cbb-b68a-41…
-
How can we use albumentations here ? or any other TOrch vision Augmentations
-
Can you explain how to use the `augment` mode in modeling inferencing?
-
使用export_onnx.py导出的onnx模型做推理,输出的shape为(1, 4, 80, 80, 19),而原模型输出的为[1, 89760, 19],请问下是什么原因造成的呢,或者有没有什么办法保持两者输出一致呢?