Android API FaceMesh inputBitmap() method crashes app

[heckedbyfurki]

I would like to get the input image when face mesh detects someone. When I try to call inputImage() method within FaceMeshResultListener I get an (Fatal signal 11 (SIGSEGV), code 1 (SEGV_MAPERR), fault addr 0x28 in tid 1948 (Thread-10), pid 1884 (le.facemeshtest)) error.

The root of the problem is probably this (Java_com_google_mediapipe_framework_PacketGetter_nativeGetGpuBuffer+376)

System information (Please provide as much relevant information as possible)

• I'm using example code provided by Mediapipe
• Android Version: 10
• Device: Xiaomi Mi9T
• Language: Java
• Version: Latest release
• Solution: FaceMesh

Describe the current behavior:
App crashes

Describe the expected behavior:
I should get a bitmap

Standalone code to reproduce the issue:
Calling inputBitmap() method

Other info / Complete Logs :
Logcat log that starts from first error message to last one

18:48:34.390 A Fatal signal 11 (SIGSEGV), code 1 (SEGV_MAPERR), fault addr 0x28 in tid 1948 (Thread-10), pid 1884 (le.facemeshtest)
18:48:34.446 D Created output texture: 381 width: 960 height: 1280
18:48:34.456 D Created output texture: 382 width: 960 height: 1280
18:48:34.460 I I20221117 18:48:34.459964 1956 jni_util.cc:41] GetEnv: not attached
18:48:34.485 D Created output texture: 383 width: 960 height: 1280
18:48:34.576 D Created output texture: 384 width: 960 height: 1280
18:48:34.582 I I20221117 18:48:34.582705 1952 jni_util.cc:41] GetEnv: not attached
18:48:34.603 A pid: 1884, tid: 1948, name: Thread-10 >>> com.example.facemeshtest <<<
18:48:34.610 D Created output texture: 385 width: 960 height: 1280
18:48:34.646 D Created output texture: 386 width: 960 height: 1280
18:48:34.684 D Created output texture: 387 width: 960 height: 1280
18:48:34.731 D Created output texture: 388 width: 960 height: 1280
18:48:34.766 D Created output texture: 389 width: 960 height: 1280
18:48:34.812 D Created output texture: 390 width: 960 height: 1280
18:48:34.852 D Created output texture: 391 width: 960 height: 1280
18:48:34.890 D Created output texture: 392 width: 960 height: 1280
18:48:34.935 E packageMonitor onPackageModified packageName=com.example.facemeshtest mIsRecentsWithinLauncher=true
18:48:34.935 D Created output texture: 393 width: 960 height: 1280
18:48:34.984 D Created output texture: 394 width: 960 height: 1280
18:48:35.065 D Created output texture: 395 width: 960 height: 1280
18:48:35.105 D Created output texture: 396 width: 960 height: 1280
18:48:35.153 D Created output texture: 397 width: 960 height: 1280
18:48:35.215 D Created output texture: 398 width: 960 height: 1280
18:48:35.260 D Created output texture: 399 width: 960 height: 1280
18:48:35.299 A #00 pc 000000000076d178 /data/app/com.example.facemeshtest-E_CRl3xBiigAA8-J1eEWqQ==/base.apk (offset 0x472000)
18:48:35.299 A #1 pc 000000000013810c /data/app/com.example.facemeshtest-E_CRl3xBiigAA8-J1eEWqQ==/base.apk (offset 0x472000) (Java_com_google_mediapipe_framework_PacketGetter_nativeGetGpuBuffer+376)
18:48:35.300 A #9 pc 0000000000a8aa9e /data/app/com.example.facemeshtest-E_CRl3xBiigAA8-J1eEWqQ==/oat/arm64/base.vdex (com.google.mediapipe.framework.PacketGetter.getTextureFrame+14)
18:48:35.300 A #12 pc 0000000000a9e8e0 /data/app/com.example.facemeshtest-E_CRl3xBiigAA8-J1eEWqQ==/oat/arm64/base.vdex (com.google.mediapipe.solutioncore.ImageSolutionResult.acquireInputTextureFrame+12)
18:48:35.300 A #15 pc 0000000000a9f214 /data/app/com.example.facemeshtest-E_CRl3xBiigAA8-J1eEWqQ==/oat/arm64/base.vdex (com.google.mediapipe.solutioncore.SolutionGlSurfaceViewRenderer.setRenderData)
18:48:35.300 A #18 pc 0000000000a9f346 /data/app/com.example.facemeshtest-E_CRl3xBiigAA8-J1eEWqQ==/oat/arm64/base.vdex (com.google.mediapipe.solutioncore.SolutionGlSurfaceView.setRenderData+6)
18:48:35.300 A #21 pc 00000000008fe340 /data/app/com.example.facemeshtest-E_CRl3xBiigAA8-J1eEWqQ==/oat/arm64/base.vdex (com.example.facemeshtest.MainActivity.lambda$setupStreamingModePipeline$2$com-example-facemeshtest-MainActivity+12)
18:48:35.300 A #24 pc 00000000008fe254 /data/app/com.example.facemeshtest-E_CRl3xBiigAA8-J1eEWqQ==/oat/arm64/base.vdex (com.example.facemeshtest.MainActivity$$ExternalSyntheticLambda2.run+8)
18:48:35.300 A #27 pc 0000000000a9ea98 /data/app/com.example.facemeshtest-E_CRl3xBiigAA8-J1eEWqQ==/oat/arm64/base.vdex (com.google.mediapipe.solutioncore.OutputHandler.run+52)
18:48:35.300 A #30 pc 0000000000a9ebe4 /data/app/com.example.facemeshtest-E_CRl3xBiigAA8-J1eEWqQ==/oat/arm64/base.vdex (com.google.mediapipe.solutioncore.SolutionBase$$ExternalSyntheticLambda0.process+4)
18:48:35.300 A #41 pc 000000000012545c /data/app/com.example.facemeshtest-E_CRl3xBiigAA8-J1eEWqQ==/base.apk (offset 0x472000)
18:48:35.300 A #42 pc 0000000000125688 /data/app/com.example.facemeshtest-E_CRl3xBiigAA8-J1eEWqQ==/base.apk (offset 0x472000)
18:48:35.300 A #43 pc 000000000073b934 /data/app/com.example.facemeshtest-E_CRl3xBiigAA8-J1eEWqQ==/base.apk (offset 0x472000)
18:48:35.300 A #44 pc 000000000075dfac /data/app/com.example.facemeshtest-E_CRl3xBiigAA8-J1eEWqQ==/base.apk (offset 0x472000)
18:48:35.300 A #45 pc 0000000000752808 /data/app/com.example.facemeshtest-E_CRl3xBiigAA8-J1eEWqQ==/base.apk (offset 0x472000)
18:48:35.300 A #46 pc 00000000007523bc /data/app/com.example.facemeshtest-E_CRl3xBiigAA8-J1eEWqQ==/base.apk (offset 0x472000)
18:48:35.300 A #47 pc 00000000007749e0 /data/app/com.example.facemeshtest-E_CRl3xBiigAA8-J1eEWqQ==/base.apk (offset 0x472000)
18:48:35.300 A #48 pc 0000000000774748 /data/app/com.example.facemeshtest-E_CRl3xBiigAA8-J1eEWqQ==/base.apk (offset 0x472000)

[Camokaliv]

Have you tried using CPU instead of GPU?

[heckedbyfurki]

Yeah, nothing happens. When I set RUN_ON_GPU to false, it crashes instantly. As I see it's a bug that has still not been solved.

[sohaiberrabii]

I have encountered a similar issue on the Java solutions API. I worked around it by calling inputBitmap after setRenderData.

[heckedbyfurki]

I have encountered a similar issue on the Java solutions API. I worked around it by calling inputBitmap after setRenderData.

This. You are a legend

[google-ml-butler]

Are you satisfied with the resolution of your issue?
Yes
No