You have to do a work-around like it was answered here in the past.
Check this test to see how it is done:
http://code.opencv.org/projects/opencv/repository/revisions/master/entry/modules/java/android_test/src/org/opencv/test/features2d/ORBDescriptorExtractorTest.java
These comments have ORB default parameters values that I saw within the source code, but these might be outdated and might have errors. Chech the source yourself to make sure.
// ORB Default values:
// float scaleFactor = 1.2f // Coefficient by which we divide the dimensions from one scale pyramid level to the next
// uint nLevels = 8 // The number of levels in the scale pyramid
// uint firstLevel = 0 // The level at which the image is given
// int edgeThreshold = 31 // How far from the boundary the points should be.
// int patchSize = 31 // You can not change this, it is allways 31
// int WTA_K = 2 // How many random points are used to produce each cell of the descriptor (2, 3, 4 ...)
// scoreType = 0 // 0 for HARRIS_SCORE / 1 for FAST_SCORE
// nFeatures = 500 // not sure if 500 is default
String filename = "/path/to/file/orb_params.yml";
writeFile(paramFile, "%YAML:1.0\nscaleFactor: 1.2\nnLevels: 8\nfirstLevel: 0 \nedgeThreshold: 31\npatchSize: 31\nWTA_K: 2\nscoreType: 0\nnFeatures: 500\n");
extractor.read(filename); //note: this will fail if the above parameters are not the ones expected by method
detector.detect(img, keypoints);
(...)