@LsyAi If you're referring to the Segment-Everything model (which segments everything in the image), it's not straightforward. You would need to export the encoder and decoder and then create a wrapper around them that does all the processing. Exporting it as a single model to ONNX wasn't possible for me and I'm not sure if it's even achievable, I've encountered many ONNX exceptions when trying to do so and was unable to fix them
@LsyAi If you're referring to the Segment-Everything model (which segments everything in the image), it's not straightforward. You would need to export the encoder and decoder and then create a wrapper around them that does all the processing. Exporting it as a single model to ONNX wasn't possible for me and I'm not sure if it's even achievable, I've encountered many ONNX exceptions when trying to do so and was unable to fix them