rscnn
rscnn copied to clipboard
Some method are not implemented
HI, I am trying to run your demo, but loadParams from BatchNorm.rs is not implemented. The method is returning a message saying: "fastLoadParams is not implemented". The parsing operation goes ahead but when it is trying to instantiate the network is crashing after conv0 This is the error message that I got: """ 2018-11-02 13:21:16.230 5543-5543/com.rscnn.example I/LayerGraph: layer MobileNet-SSD output:,1,300,300,3 2018-11-02 13:21:16.233 5543-5543/com.rscnn.example I/LayerGraph: layer data output:,1,300,300,3 2018-11-02 13:21:16.248 5543-5543/com.rscnn.example I/LayerGraph: layer conv0 output:,1,150,150,32
--------- beginning of crash
2018-11-02 13:21:16.252 5543-5543/com.rscnn.example E/AndroidRuntime: FATAL EXCEPTION: main
Process: com.rscnn.example, PID: 5543
java.lang.RuntimeException: Unable to start activity ComponentInfo{com.rscnn.example/com.rscnn.example.activity.MainActivity}: java.lang.NullPointerException: Attempt to get length of null array
at android.app.ActivityThread.performLaunchActivity(ActivityThread.java:2896)
at android.app.ActivityThread.handleLaunchActivity(ActivityThread.java:2974)
at android.app.ActivityThread.-wrap11(Unknown Source:0)
at android.app.ActivityThread$H.handleMessage(ActivityThread.java:1663)
at android.os.Handler.dispatchMessage(Handler.java:108)
at android.os.Looper.loop(Looper.java:206)
at android.app.ActivityThread.main(ActivityThread.java:6749)
at java.lang.reflect.Method.invoke(Native Method)
at com.android.internal.os.Zygote$MethodAndArgsCaller.run(Zygote.java:240)
at com.android.internal.os.ZygoteInit.main(ZygoteInit.java:845)
Caused by: java.lang.NullPointerException: Attempt to get length of null array
at com.rscnn.layers.BatchNorm.setup(BatchNorm.java:56)
at com.rscnn.network.LayerGraph.init(LayerGraph.java:294)
at com.rscnn.network.LayerGraph.init(LayerGraph.java:326)
at com.rscnn.network.LayerGraph.init(LayerGraph.java:326)
at com.rscnn.network.LayerGraph.init(LayerGraph.java:326)
at com.rscnn.network.ConvNet.loadGraphFromRawData(ConvNet.java:96)
at com.rscnn.network.ConvNet.
It looks to me that is crashing when trying to instantiate the batch normalization layer after conv0.
COuld you clarify please?
Thank you very much
@contactmat85 You should merge the batch norm layers into conv layers, use my merge_bn.py in mobilenet-ssd project to do that.