-
### Description
I tried to fix a build error of [ncnn](https://github.com/Tencent/ncnn) for Android 'arm64-v8a' ABI with NDK r27 by enabling cmake policy CMP0057 for cmake version >=3.3 in PR https…
-
I tried creating a convert function in JNI (which calls the process function of RealSR) and used it in Kotlin. However, the output image after upscaling is just black. I would appreciate your advice o…
-
已经参考 [Convert yolov8 model to ncnn model · Digital2Slave/ncnn-android-yolov8-seg Wiki (github.com)](https://github.com/Digital2Slave/ncnn-android-yolov8-seg/wiki/Convert-yolov8-model-to-ncnn-model)
…
-
In file included from E:/download/android-yolo/ByteTrack_ncnn_Android-main/app/src/main/jni/opencv-4.5.1-android-sdk/sdk/native/jni/include\opencv2/flann/params.h:35:
E:/download/android-yolo/ByteTra…
-
数字人 DEMO 在 RK3588 板子上运行流畅,wav驱动数字人说话的时候单帧渲染在 40毫秒以内。但在 RK3568 (4 核心A55 , 4GB 内存)上,说话的时候卡顿很厉害,单帧渲染去到 150毫秒左右。4核心的 CPU 配置在硅基数字人的说明中也是支持的,但卡的这么厉害。
由于我需要在 4核 A55 的环境里运用硅基数字人,所以只能进行优化,目前已经将说话时候的单帧渲染优化到 10…
-
Please see
https://github.com/Tencent/ncnn/actions/workflows/android.yml
-
# Why is my 1D convolution implementation not working?
The output shape comes out to 0x0x0.
## Also where is the C API documenation?
More info:
I am linking against the arm64-v8a version for a…
-
I am having trouble understanding your code, specifically at line 137 in the file waifu2x_jni.cpp (waifu2x->process(in, out);). It seems that after the model processes, the ncnn::Mat out is not being …
-
Thanks for your amazing project sherpa-ncnn and it provide an alternative excellent ASR engine to open source community and developers.
Could you implement TTS using next-generation Kaldi with [ncn…
-
我使用的模型是`sherpa-ncnn-streaming-zipformer-zh-14M-2023-02-23`,目前想使用`x86设备的N卡`做推理,想问问这个模型是否支持GPU以及对应的操作办法。另外想咨询一下这个模型是否支持`边缘计算设备上的NPU`,比如RK系列开发板。