FeiYull
Yolov8 classification
Hello,this is an excellent job! I have a question here,how can I implement yolov8-cls?
yolov8-cls +1
You only need to remove its post-processing operation on yolov8-det, other than that, other operations do not need to change
yolov8-cls +1
You only need to remove its post-processing operation on yolov8-det, other than that, other operations do not need to change
大佬你好，那推理的结果我要怎么使用呢，不是很清楚谢谢
yolov8-cls +1
You only need to remove its post-processing operation on yolov8-det, other than that, other operations do not need to change
大佬你好，那推理的结果我要怎么使用呢，不是很清楚谢谢
@dabensongbing 你好，处理结果后使用std::max_element()找出最高的分数就好了
yolov8-cls +1
You only need to remove its post-processing operation on yolov8-det, other than that, other operations do not need to change
大佬你好，那推理的结果我要怎么使用呢，不是很清楚谢谢
@dabensongbing 你好，处理结果后使用std::max_element()找出最高的分数就好了
非常感谢，我的主要问题是拷贝数据时可能自己之前代码不对，现在已经解决了，再次感谢你的回复^^
std::vector<float> outputData(param.batch_size * param.num_class); cudaMemcpy(outputData.data(), yolo.m_output_src_device, outputData.size() * sizeof(float), cudaMemcpyDeviceToHost);
