EXE_BAD_ACCESS when trying to access ObjectDetect.detector(options)

TensorFlow General Discussion
, ,
1

When trying to access to ObjectDetect.detector(options). Getting bad memory access error.


    init?(modelFileInfo: FileInfo, threadCount: Int, scoreThreshold: Float, maxResults: Int) {
        let modelFilename = modelFileInfo.name
        
        // Construct the path to the model file.
        guard let modelPath = Bundle.main.path(
                forResource: modelFileInfo.name,
                ofType: modelFileInfo.extension
            )
        else {
            print("Failed to load the model file with name: \(modelFilename).")
            return nil
        }
        
        // Specify the options for the `Detector`.
        let options = ObjectDetectorOptions(modelPath: modelPath)
        options.classificationOptions.scoreThreshold = scoreThreshold
        options.classificationOptions.maxResults = maxResults
        options.baseOptions.computeSettings.cpuSettings.numThreads = threadCount
        do {
            // Create the `Detector`.
            detector = try ObjectDetector.detector(options: options)
        } catch let error {
            print("Failed to create the interpreter with error: \(error.localizedDescription)")
            return nil
        }
        
        super.init()
    }

i got this code from examples/lite/examples/object_detection/ios at master ยท tensorflow/examples (github.com)

when i run this project cloned from the above git its working perfectly.

to re-produce the error:
create new ios project (i am using ios 17.2) and import the library then try to process a image. it will produce the EXE_BAD_ACCESS when trying to

detector = try ObjectDetector.detector(options: options)