- `NonMaxSuppression`
- Workaround for NonMaxSuppression bug in TensorFlow Lite. Migration to `NonMaxSuppressionV2`
- However, all NMS will be FlexOP. This is a silly fix, I know, but it is unavoidable to avoid output inconsistencies.
- https://github.com/tensorflow/tensorflow/issues/51629
- https://github.com/tensorflow/tensorflow/issues/52388
- [Bug in TFlite NonMaxSuppression 48](https://github.com/PINTO0309/onnx2tf/issues/48)
What's Changed
* [experimental] https://github.com/tensorflow/tensorflow/issues/51629 by PINTO0309 in https://github.com/PINTO0309/onnx2tf/pull/49
**Full Changelog**: https://github.com/PINTO0309/onnx2tf/compare/1.2.9...1.2.10