topic Re: imx93 dms face_landmark model, the output of the model is not consistent with the output process in i.MX Processors

imx93 dms face_landmark model, the output of the model is not consistent with the output processing

DOMINICK — Mon, 08 Jan 2024 06:45:23 GMT

The face_landmark model downloaded by download_models.py in /usr/bin/eiq-examples-git of the im93 sdk version 6.1.55 is inconsistent with the output processing of the routine. If possible, please provide a model that fits this routine

Re: imx93 dms face_landmark model, the output of the model is not consistent with the output process

Zhiming_Liu — Tue, 09 Jan 2024 05:51:43 GMT

The url is not maintained by NXP, you can refer this github link:

https://github.com/PINTO0309/PINTO_model_zoo/blob/main/043_face_landmark/download.sh

Re: imx93 dms face_landmark model, the output of the model is not consistent with the output process

DOMINICK — Fri, 12 Jan 2024 02:46:49 GMT

你好，我明白不是nxp维护的这个模型，只是现在imx93里面该目录下的推理例程和下载下来的这个模型不匹配了（可能这个网站做了模型迭代导致的），麻烦反馈提供新的模型或者新的推理例程或者修改方法，现在是模型的输出和例程输出完全不匹配直接会导致推理错误。

Re: imx93 dms face_landmark model, the output of the model is not consistent with the output process

Zhiming_Liu — Fri, 12 Jan 2024 05:31:32 GMT

你用的是什么命令测的？CPU推理还是NPU?

Re: imx93 dms face_landmark model, the output of the model is not consistent with the output process

DOMINICK — Fri, 12 Jan 2024 06:44:35 GMT

1、使用的是cpu推理，执行命令python3 main.py -i /dev/video1 (我的usb摄像头是video1)

2、使用下载的landmark的模型根据下载脚本提供解压后的resources\10_new_detector_landmark\resources\saved_model里面的model_float32.tflite，我先把模型名字改成了和例程一致的，然后发现输出实际不一致，通过netron查看，发现和例程要求的名字不一样，例程要求的是2输出（可以看face_landmark.py里面），但实际模型是3输出

麻烦实际测一下就知道问题所在了

Re: imx93 dms face_landmark model, the output of the model is not consistent with the output process

Zhiming_Liu — Fri, 12 Jan 2024 07:42:23 GMT

你用这个里面的模型，不要用新的。

Re: imx93 dms face_landmark model, the output of the model is not consistent with the output process

DOMINICK — Fri, 12 Jan 2024 08:14:47 GMT

可以用了，谢谢