[Bug] detction result from mmdeploy model are different from openmmlab

[lileikang1]

Checklist

• I have searched related issues but cannot get the expected help.
• 2. I have read the FAQ documentation but cannot get the expected help.
• 3. The bug has not been fixed in the latest version.

Describe the bug

detction result from mmdeploy model are different from openmmlab
model is faster-rcnn
[图片]
[图片]
[图片]
[图片]

Reproduction

1
./build/object_detection cuda mmdeploy_model/faster_rcnn_pipe_fp32 data/test_pipe
2
yes, my changes wouldn't affect the results of detection.

Environment

1
2023-01-31 17:59:16,221 - mmdeploy - INFO - **********Environmental information**********
2023-01-31 17:59:16,457 - mmdeploy - INFO - sys.platform: linux
2023-01-31 17:59:16,457 - mmdeploy - INFO - Python: 3.9.15 (main, Nov 24 2022, 14:31:59) [GCC 11.2.0]
2023-01-31 17:59:16,457 - mmdeploy - INFO - CUDA available: True
2023-01-31 17:59:16,457 - mmdeploy - INFO - GPU 0: GeForce GTX 1660 SUPER
2023-01-31 17:59:16,457 - mmdeploy - INFO - CUDA_HOME: /usr/local/cuda-11.0
2023-01-31 17:59:16,457 - mmdeploy - INFO - NVCC: Cuda compilation tools, release 11.0, V11.0.194
2023-01-31 17:59:16,457 - mmdeploy - INFO - GCC: gcc (Ubuntu 9.4.0-1ubuntu1~16.04) 9.4.0
2023-01-31 17:59:16,457 - mmdeploy - INFO - PyTorch: 1.9.1+cu111
2023-01-31 17:59:16,457 - mmdeploy - INFO - PyTorch compiling details: PyTorch built with:
  - GCC 7.3
  - C++ Version: 201402
  - Intel(R) oneAPI Math Kernel Library Version 2021.4-Product Build 20210904 for Intel(R) 64 architecture applications
  - Intel(R) MKL-DNN v2.1.2 (Git Hash 98be7e8afa711dc9b66c8ff3504129cb82013cdb)
  - OpenMP 201511 (a.k.a. OpenMP 4.5)
  - NNPACK is enabled
  - CPU capability usage: AVX2
  - CUDA Runtime 11.1
  - NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86
  - CuDNN 8.0.5
  - Magma 2.5.2
  - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.1, CUDNN_VERSION=8.0.5, CXX_COMPILER=/opt/rh/devtoolset-7/root/usr/bin/c++, CXX_FLAGS= -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -fopenmp -DNDEBUG -DUSE_KINETO -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -O2 -fPIC -Wno-narrowing -Wall -Wextra -Werror=return-type -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-sign-compare -Wno-unused-parameter -Wno-unused-variable -Wno-unused-function -Wno-unused-result -Wno-unused-local-typedefs -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=1.9.1, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, 

2023-01-31 17:59:16,457 - mmdeploy - INFO - TorchVision: 0.10.1+cu111
2023-01-31 17:59:16,457 - mmdeploy - INFO - OpenCV: 4.6.0
2023-01-31 17:59:16,457 - mmdeploy - INFO - MMCV: 1.7.0
2023-01-31 17:59:16,458 - mmdeploy - INFO - MMCV Compiler: GCC 7.3
2023-01-31 17:59:16,458 - mmdeploy - INFO - MMCV CUDA Compiler: 11.1
2023-01-31 17:59:16,458 - mmdeploy - INFO - MMDeploy: 0.11.0+2a1fed9
2023-01-31 17:59:16,458 - mmdeploy - INFO - 

2023-01-31 17:59:16,458 - mmdeploy - INFO - **********Backend information**********
2023-01-31 17:59:17,040 - mmdeploy - INFO - onnxruntime: 1.12.1 ops_is_avaliable : False
2023-01-31 17:59:17,061 - mmdeploy - INFO - tensorrt: 8.2.3.0   ops_is_avaliable : True
2023-01-31 17:59:17,084 - mmdeploy - INFO - ncnn: None  ops_is_avaliable : False
2023-01-31 17:59:17,089 - mmdeploy - INFO - pplnn_is_avaliable: False
2023-01-31 17:59:17,096 - mmdeploy - INFO - openvino_is_avaliable: False
2023-01-31 17:59:17,121 - mmdeploy - INFO - snpe_is_available: False
2023-01-31 17:59:17,129 - mmdeploy - INFO - ascend_is_available: False
2023-01-31 17:59:17,134 - mmdeploy - INFO - coreml_is_available: False
2023-01-31 17:59:17,134 - mmdeploy - INFO - 

2023-01-31 17:59:17,134 - mmdeploy - INFO - **********Codebase information**********
2023-01-31 17:59:17,136 - mmdeploy - INFO - mmdet:      2.26.0
2023-01-31 17:59:17,136 - mmdeploy - INFO - mmseg:      None
2023-01-31 17:59:17,136 - mmdeploy - INFO - mmcls:      0.23.2
2023-01-31 17:59:17,136 - mmdeploy - INFO - mmocr:      None
2023-01-31 17:59:17,136 - mmdeploy - INFO - mmedit:     None
2023-01-31 17:59:17,136 - mmdeploy - INFO - mmdet3d:    None
2023-01-31 17:59:17,136 - mmdeploy - INFO - mmpose:     None
2023-01-31 17:59:17,136 - mmdeploy - INFO - mmrotate:   None
2023-01-31 17:59:17,136 - mmdeploy - INFO - mmaction:   None


2
tensorRT8.2.3
configs/mmdet/detection/detection_tensorrt_dynamic-320x320-1344x1344.py

Error traceback

No response