TensorRT uff parser register_input fails on 4D or higher input shape

Accelerated Computing CUDA CUDA Programming and Performance
1

I am seeing the following error from uffparser.register_input when the input is 4D or higher, e.g., (1, 356, 337, 3). How can I get around this error?

SystemError: <built-in method register_input of nv_uff_parser_bindings.UffParser object at 0x7feed1825998> returned a result with an error set

2

I do have the same issue. “TypeError: Tuple needs to have dimension 3”