any ideas on following issues

[billmguo]

Android crash when inference with mobilenet ssd.
hread #15, name = 'inference', stop reason = signal SIGSEGV: invalid address (fault address: 0x0)

• frame #0: 0x00000000
  frame How to use this? Any Doc or Samples? #1: 0xca0da56c libncnn_jni.soncnn::Net::forward_layer(int, std::__ndk1::vector<ncnn::Mat, std::__ndk1::allocator<ncnn::Mat> >&, ncnn::Option&) const [inlined] ncnn::Mat::release() at mat.h:636 frame #2: 0xca0da542 libncnn_jni.soncnn::Net::forward_layer(int, std::__ndk1::vector<ncnn::Mat, std::__ndk1::allocatorncnn::Mat >&, ncnn::Option&) const [inlined] ncnn::Mat::~Mat() at mat.h:262
  frame Added the execute permission to the shell scrips. #3: 0xca0da542 libncnn_jni.soncnn::Net::forward_layer(this=0x00000067, layer_index=<unavailable>, blob_mats=size=1, opt=0xcab0146c) const at net.cpp:884 frame #4: 0xca0d9c4a libncnn_jni.soncnn::Net::forward_layer(this=0xca1f7a24, layer_index=, blob_mats=size=1, opt=0xcab0146c) const at net.cpp:896
  frame 后续是否会支持将 tensorflow pb模型转为ncnn的模型呢? #5: 0xca0d9c4a libncnn_jni.soncnn::Net::forward_layer(this=0xca1f7a24, layer_index=<unavailable>, blob_mats=size=1, opt=0xcab0146c) const at net.cpp:896 frame #6: 0xca0da818 libncnn_jni.soncnn::Extractor::extract(this=0xcab0145c, blob_index=, feat=0xcab01438) at net.cpp:1016
  frame Update README.md #7: 0xca0cc92c libncnn_jni.so::Java_com_davidchiu_ncnncam_Ncnn_nativeDetect(env=0xcacea540, thiz=0xcab0152c, bitmap=0xcab01530) at ncnn_jni.cpp:194 frame #8: 0xcb25c60e base.odexnativeDetect(this=, (null)=) + 94
  frame How to run the example project? #9: 0xe5ddcd76 libart.soart_quick_invoke_stub_internal + 70 frame #10: 0xe5de1eea libart.soart_quick_invoke_stub + 234

[billmguo]

solved, it proves model and ncnn lib does not match