Digital2Slave

@chinmayanandValtech you need to create your custom jni function in **yolov8ncnn.cpp** like the following code snippet: ```java JNIEXPORT jobject JNICALL Java_cn_epbox_cv_1segx_CvSegx_getPicSegmentResult( JNIEnv *env, jobject p_this, jobject assetManager, jobject bmp, jstring...

> 你好，大神！yolov8-obb模型转ncnn模型，需要修改 head.py 和 block.py的那些地方，才能转换成功？谢谢！！！ > > 期待您的回复！！！ https://docs.ultralytics.com/tasks/obb/#export

1. refine your custom dataset label without missing. 2. fine tuning your segment model with your custom dataset. 3. set `prob_threshold` and `nms_threshold ` values according to your measurements.

Dear BlakeZim, Thanks for your issue. I'm very appreciate it. Please feel free to open a pull request.

**num_neighbors** is just the radius of a sliding window. Usually, according to **Imagescale** value, a sliding window is a rectangle window, which may be `5*5`, `3*3`, etc.

If **num_neighbors = Imagescale/2** is work correctly. When **Imagescale = 2, num_neighbors = 1**, and the sliding window may be smaller than `3*3` which may not work. In my code...

For your own model, please check out the load parameter of model. And refer https://github.com/Digital2Slave/ncnn-android-yolov8-seg/wiki/Convert-yolov8%E2%80%90seg-to-ncnn-model-step-by-step ``` yolov8.load_param("models/yolov8s-seg.param"); yolov8.load_model("models/yolov8s-seg.bin"); ```

what's the error message of logcat ?

``` 2023-10-12 02:06:57.371 15240-15260 OpenGLRenderer com.tencent.yolov8ncnn E fbcNotifyFrameComplete error: undefined symbol: fbcNotifyFrameComplete 2023-10-12 02:06:57.371 15240-15260 OpenGLRenderer com.tencent.yolov8ncnn E fbcNotifyNoRender error: undefined symbol: fbcNotifyNoRender ``` Refer: - https://github.com/jonbhanson/flutter_native_splash/issues/478 - https://stackoverflow.com/questions/70882079/android-app-stop-immediately-when-running-on-android-11-only-without-logcat-mess

@wsurer 最新版本的话，还是用Method 2可以的。不过，涉及的JNI代码，我还没同步。你可以参考[Method 2](https://github.com/Digital2Slave/ncnn-android-yolov8-seg/wiki/Convert-yolov8%E2%80%90seg-to-ncnn-model-step-by-step) ``` ncnn::Extractor ex = yolov8.create_extractor(); ex.input("images", in_pad); ncnn::Mat out; ex.extract("output0", out); ncnn::Mat mask_proto; ex.extract("output1", mask_proto); ```