TensorRT Conversion

Hello, I need help in converting our mtcnn face detection model through TensorRT.

Hi,

We don’t have an official example for MTCNN detector.
But below is an experiment sharing from our user:
https://jkjung-avt.github.io/tensorrt-mtcnn/

Thanks.

Thanks, this works but how do I combine this with transfer learning tool to save images and .then retain them for future use under TensorRT

Hi,

The GitHub will generate the TensorRT engine.
So you can use the file with pure TensorRT API directly.

However, TLT only support the predefined models so it cannot be used with the MTCNN shared above.
Thanks.