CenterNet精度量化效果很差

只看该作者 · 发表于 2020-1-2 16:42:26

本帖最后由 LSC 于 2020-1-2 16:47 编辑

jefferyzhang 发表于 2020-1-2 12:01
从图是看不出精度问题的，只能推断出很可能是精度问题。。。
检不捡的出还得看你NMS阈值设置多少 ...

版主大大，再请教两个问题：
1. 如何进行通道级量化？是将cfg文件中各项method参数由layer改成channel吗？
2. config file的作用是什么？

...

I Config File "/home/lsc/anaconda3/envs/RKNN/lib/python3.6/site-packages/rknn/base/RK1808_PID0X82" load/generated successfully

I Initialzing network optimizer by /home/lsc/anaconda3/envs/RKNN/lib/python3.6/site-packages/rknn/base/RK1808_PID0X82 ...

...

复制代码

只看该作者 · 发表于 2020-1-2 16:46:44

per_channel的量化我们不支持

只看该作者 · 发表于 2020-1-3 10:18:07

混合量化，将一些层从cfg文件中去掉，方法改成delete，生成模型之后运行会报这个错误。



Create RKNN model fail!

E Catch exception when init runtime!

E Traceback (most recent call last):

E   File "rknn/api/rknn_base.py", line 788, in rknn.api.rknn_base.RKNNBase.init_runtime

E   File "rknn/api/rknn_runtime.py", line 303, in rknn.api.rknn_runtime.RKNNRuntime.build_graph

E Exception: RKNN init failed. error code: RKNN_ERR_DEVICE_UNAVAILABLE

W We need do some initial work, it will increase call time of inference for the first time.

Create RKNN model fail!

E Catch exception when init model!

E Traceback (most recent call last):

E   File "rknn/api/rknn_base.py", line 901, in rknn.api.rknn_base.RKNNBase.inference

E   File "rknn/api/rknn_runtime.py", line 303, in rknn.api.rknn_runtime.RKNNRuntime.build_graph

E Exception: RKNN init failed. error code: RKNN_ERR_DEVICE_UNAVAILABLE

Traceback (most recent call last):

  File "ctdet.py", line 10, in <module>

    det.run(img)

  File "/home/lsc/rknn-toolkit-v1.3.0/centernet/ctdet_rknn_v2/detector.py", line 93, in run

    output, dets = self.process(images)

  File "/home/lsc/rknn-toolkit-v1.3.0/centernet/ctdet_rknn_v2/detector.py", line 44, in process

    hm = output[3].reshape((1,80,128,128))

TypeError: 'NoneType' object is not subscriptable
复制代码