SuperPoint 连板推理的时候出错：E Get outputs failed, error code: RKNN_ERR_FAIL

[Jiaxin0630]

想尝试 Superpoint + Lightglue 在 RK3588 上推理，在 PC 上连板推理的时候报错，不太清楚问题出在哪，目前是把 Superpoint + Lightglue 拆开单独尝试，但是尝试 Superpoint 的时候就遇到这个问题了，在PC上模拟器推理没有问题，但是连板推理就有这个问题。

I Get hardware info: target_platform = rk3588, os = Android, aarch = aarch64
I Check RK3588 board npu runtime version
I Starting ntp or adb, target is RK3588
I Start adb...
I Connect to Device success!
I NPUTransfer(880891): Starting NPU Transfer Client, Transfer version 2.2.2 (12abf2a@2024-09-02T03:22:41)
D NPUTransfer(880891): Transfer spec = local:transfer_proxy
D NPUTransfer(880891): Transfer interface successfully opened, fd = 3
I NPUTransfer(880891): TransferBuffer: min aligned size: 1024
D RKNNAPI: ==============================================
D RKNNAPI: RKNN VERSION:
D RKNNAPI: API: 2.3.2 (1842325 build@2025-03-30T09:55:23)
D RKNNAPI: DRV: rknn_server: 2.3.2 (1842325 build@2025-03-30T09:54:16)
D RKNNAPI: DRV: rknnrt: 2.3.2 (429f97ae6b@2025-04-09T09:08:16)
D RKNNAPI: ==============================================
D RKNNAPI: Input tensors:
D RKNNAPI: index=0, name=images, n_dims=4, dims=[1, 480, 640, 1], n_elems=307200, size=614400, w_stride = 0, size_with_stride = 0, fmt=NHWC, type=FP16, qnt_type=NONE, zp=0, scale=1.000000
D RKNNAPI: Output tensors:
D RKNNAPI: index=0, name=keypoints, n_dims=3, dims=[1, 2048, 2], n_elems=4096, size=8192, w_stride = 0, size_with_stride = 0, fmt=UNDEFINED, type=FP16, qnt_type=NONE, zp=0, scale=1.000000
D RKNNAPI: index=1, name=scores, n_dims=2, dims=[1, 2048], n_elems=2048, size=4096, w_stride = 0, size_with_stride = 0, fmt=UNDEFINED, type=FP16, qnt_type=NONE, zp=0, scale=1.000000
D RKNNAPI: index=2, name=descriptors, n_dims=3, dims=[1, 2048, 256], n_elems=524288, size=1048576, w_stride = 0, size_with_stride = 0, fmt=UNDEFINED, type=FP16, qnt_type=NONE, zp=0, scale=1.000000
done
--> Preprocessing images
--> Testing superpoint inference
E RKNNAPI: rknn_outputs_get, msg_finish fail, result = 1(ACK_FAIL), expect 0(ACK_SUCC)!!
E inference: Traceback (most recent call last):
File "rknn/api/rknn_log.py", line 344, in rknn.api.rknn_log.error_catch_decorator.error_catch_wrapper
File "rknn/api/rknn_base.py", line 2779, in rknn.api.rknn_base.RKNNBase.inference
File "rknn/api/rknn_runtime.py", line 613, in rknn.api.rknn_runtime.RKNNRuntime.get_outputs
Exception: E Get outputs failed, error code: RKNN_ERR_FAIL

W inference: ===================== WARN(1) =====================
E rknn-toolkit2 version: 2.3.2
Traceback (most recent call last):
File "rknn/api/rknn_log.py", line 344, in rknn.api.rknn_log.error_catch_decorator.error_catch_wrapper
File "rknn/api/rknn_base.py", line 2779, in rknn.api.rknn_base.RKNNBase.inference
File "rknn/api/rknn_runtime.py", line 613, in rknn.api.rknn_runtime.RKNNRuntime.get_outputs
Exception: E Get outputs failed, error code: RKNN_ERR_FAIL

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "extract_superpoint_rknn.py", line 135, in
batch_outputs = rknn.inference(inputs=[img1], data_format=['nhwc'])
File "/home/jiaxin/anaconda3/envs/rk/lib/python3.8/site-packages/rknn/api/rknn.py", line 314, in inference
return self.rknn_base.inference(inputs=inputs, data_format=data_format,
File "rknn/api/rknn_log.py", line 349, in rknn.api.rknn_log.error_catch_decorator.error_catch_wrapper
File "rknn/api/rknn_log.py", line 95, in rknn.api.rknn_log.RKNNLog.e
ValueError: Traceback (most recent call last):
File "rknn/api/rknn_log.py", line 344, in rknn.api.rknn_log.error_catch_decorator.error_catch_wrapper
File "rknn/api/rknn_base.py", line 2779, in rknn.api.rknn_base.RKNNBase.inference
File "rknn/api/rknn_runtime.py", line 613, in rknn.api.rknn_runtime.RKNNRuntime.get_outputs
Exception: E Get outputs failed, error code: RKNN_ERR_FAIL

[lrm2017]

朋友解决了么

[xcao0045]

@Jiaxin0630 大佬，我最近也想尝试 Superpoint + Lightglue 在 RK3588 上推理，请问你解决了吗

[xiehousen]

Lightglue应该不支持吧