StarDist 2D bugging after latest updates of ImageJ

[ctrueden]

In imagej/ImageJ#267, @mpalaibmb wrote:

Good morning everyone!
This morning, while initializing ImageJ, it proposed me numerous updates that I accepted as usual.
The problem is that after them, StarDist 2D plugin, which is the one I use mostly for my analysis, started to bug.
I attach following the type of error that it gives me, which before today I never had:

[INFO] No TF library found in C:\Users\mpabm\DOWNLO~1\FIJI-W~1\Fiji.app\lib\.
[INFO] Using default TensorFlow version from JAR: TF 1.15.0 CPU
[INFO] Loading TensorFlow model GenericNetwork_aea3be563cb56b8824f53a8c2382aaa5 from source file file:/C:/Users/mpabm/AppData/Local/Temp/stardist_model_8155415352555340585.zip
[INFO] Caching TensorFlow models to C:\Users\mpabm\DOWNLO~1\FIJI-W~1\Fiji.app\models
java.util.concurrent.ExecutionException: java.lang.NoSuchMethodError: org.tensorflow.framework.MetaGraphDef.makeExtensionsImmutable()V
	at java.util.concurrent.FutureTask.report(FutureTask.java:122)
	at java.util.concurrent.FutureTask.get(FutureTask.java:192)
	at de.csbdresden.csbdeep.commands.GenericCoreNetwork.run(GenericCoreNetwork.java:420)
	at org.scijava.command.CommandModule.run(CommandModule.java:196)
	at org.scijava.module.ModuleRunner.run(ModuleRunner.java:165)
	at org.scijava.module.ModuleRunner.call(ModuleRunner.java:125)
	at org.scijava.module.ModuleRunner.call(ModuleRunner.java:64)
	at org.scijava.thread.DefaultThreadService.lambda$wrap$2(DefaultThreadService.java:247)
	at java.util.concurrent.FutureTask.run(FutureTask.java:266)
	at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
	at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
	at java.lang.Thread.run(Thread.java:750)
Caused by: java.lang.NoSuchMethodError: org.tensorflow.framework.MetaGraphDef.makeExtensionsImmutable()V
	at org.tensorflow.framework.MetaGraphDef.<init>(MetaGraphDef.java:168)
	at org.tensorflow.framework.MetaGraphDef.<init>(MetaGraphDef.java:25)
	at org.tensorflow.framework.MetaGraphDef$1.parsePartialFrom(MetaGraphDef.java:4257)
	at org.tensorflow.framework.MetaGraphDef$1.parsePartialFrom(MetaGraphDef.java:4252)
	at com.google.protobuf.AbstractParser.parsePartialFrom(AbstractParser.java:135)
	at com.google.protobuf.AbstractParser.parseFrom(AbstractParser.java:168)
	at com.google.protobuf.AbstractParser.parseFrom(AbstractParser.java:180)
	at com.google.protobuf.AbstractParser.parseFrom(AbstractParser.java:185)
	at com.google.protobuf.AbstractParser.parseFrom(AbstractParser.java:25)
	at org.tensorflow.framework.MetaGraphDef.parseFrom(MetaGraphDef.java:2680)
	at de.csbdresden.csbdeep.network.model.tensorflow.TensorFlowNetwork.loadModel(TensorFlowNetwork.java:174)
	at de.csbdresden.csbdeep.network.model.DefaultNetwork.loadModel(DefaultNetwork.java:76)
	at de.csbdresden.csbdeep.network.DefaultModelLoader.loadNetwork(DefaultModelLoader.java:69)
	at de.csbdresden.csbdeep.network.DefaultModelLoader.run(DefaultModelLoader.java:48)
	at de.csbdresden.csbdeep.commands.GenericCoreNetwork.tryToPrepareInputAndNetwork(GenericCoreNetwork.java:524)
	at de.csbdresden.csbdeep.commands.GenericCoreNetwork.initiateModelIfNeeded(GenericCoreNetwork.java:303)
	at de.csbdresden.csbdeep.commands.GenericCoreNetwork.mainThread(GenericCoreNetwork.java:445)
	at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
	... 4 more
[INFO] CSBDeep plugin exit (took 861 milliseconds)
[ERROR] Command errored: StarDist 2D
java.lang.NullPointerException
	at de.csbdresden.stardist.StarDist2D.splitPrediction(StarDist2D.java:338)
	at de.csbdresden.stardist.StarDist2D.run(StarDist2D.java:307)
	at org.scijava.command.CommandModule.run(CommandModule.java:196)
	at org.scijava.module.ModuleRunner.run(ModuleRunner.java:165)
	at org.scijava.module.ModuleRunner.call(ModuleRunner.java:125)
	at org.scijava.module.ModuleRunner.call(ModuleRunner.java:64)
	at org.scijava.thread.DefaultThreadService.lambda$wrap$2(DefaultThreadService.java:247)
	at java.util.concurrent.FutureTask.run(FutureTask.java:266)
	at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
	at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
	at java.lang.Thread.run(Thread.java:750)

I tried to update TF models, used them through a GPU, and download new packages and versions of CUDA and cuDNN.
But nothing changed. Maybe I am missing the point.

Thank you in advance for whoever can provide me with insights!

[ctrueden]

I am able to reproduce this error on my Ubuntu Linux 24.04 system with a Fiji that has the CSBDeep and StarDist update sites freshly enabled.

I tried Edit → Options → TensorFlow... to swap the versions around between 1.15 and 1.12, both CPU and GPU, but still receive the same error regardless.

@uschmidt83 Any idea what could have changed?

[mpalaibmb]

@ctrueden I actually resolved my problem changing to 1.12 version.
I saw that many people with Windows as me are having this error, so maybe since you have Linux can be different!

[ctrueden]

@mpalaibmb Glad it is working for you now! I wonder though if there is something we could fix on the CSBDeep and/or StarDist update sites so that it works out of the box...

[uschmidt83]

All the TensorFlow stuff is being handled by the CSBDeep plugin, which is unfortunately unmaintained. It's some sort of miracle that it still works to this day, to be honest.

Similar issues have been discussed in the image.sc forum, including me being confused on this topic.

Sorry that I don't have a better answer for you at the moment.

[sjcross]

I've been having a look at this and believe it's caused by the recent (2025-01-25) update of protobuf from version 3.23.0 to 4.28.2 on the Java-8 update site. On my Windows machine taking a fresh copy of Fiji with just the Stardist and CSBDeep update sites added, then downgrading protobuf back to v3.23.0 fixed the problem.

This also explains why adding the Tensorflow update site seems to be fixing the problem, as this comes with an old protobuf (v3.2.0). Interestingly, the CSBDeep site comes with protobuf v3.3.1, but unlike the Tensorflow site, this doesn't appear to override the newer protobuf coming from Java-8.

From the little I can find online, makeExtensionsImmutable() does indeed appear to have been removed in recent versions of protobuf (there's mention of it in this thread https://lists.apache.org/thread/wmtokql9lrtks28zv5d35mbz3g0ct5d6).

[loicsauteur]

Thanks for these useful comments.

On macOS (apple Silicon) I could also get the StarDist plugin to work, steps:

1. Download a fresh copy of Fiji (Intel version)
2. Install following update sites:
   • CSBDeep
   • StarDist
   • Tensorflow

Doing this with the Fiji macOS arm64 version did not work. There I got "Could not load TensorFlow. Opening the TensorFlow Library Management tool"

Cannot find TensorFlow native library for OS: darwin, architecture: aarch64. 
See https://github.com/tensorflow/tensorflow/tree/master/tensorflow/java/README.md 
for possible solutions (such as building the library from source). 
Additional information on attempts to find the native library can be obtained by 
adding org.tensorflow.NativeLibrary.DEBUG=1 to the system properties of the JVM.