Challenges with YOLOv5_custom data Implementation on Jetson Nano Orin: Limited Label Detection and Bounding Box Issues

[2509afroz]

@marcoslucianops Hello, I've employed YOLOv5 for training on a GPU using custom data, then exported it to obtain a .onnx file. I aimed to implement this workflow on a Jetson Nano Orin for inference. However, during inference, I encountered an issue where only one label is displayed from the labels.txt file, while other objects are not detected. Moreover, the bounding boxes seem to encapsulate the entire image, resembling a single object detection, such as a person bounding box.