-
MacOS平台下使用MNN做背景替换
使用的是lite.ai.toolkit中的代码，网址如下：
https://github.com/DefTruth/lite.ai.toolkit
当不使用Metal的时候是正常的，但是如果开启了Metal，效果会非常差
`
#include "mnn_rvm.h"
#include "lite/utils.h"
using mnncv:…
-
Hope ssdlite-mobilenetV3-small can be used under tensorflow 2.x.x.I hope the official can realize it soon.
-
针对文档[PP-OCRv3 文本检测模型训练](https://github.com/PaddlePaddle/PaddleOCR/blob/18ddb6d5f9bdc2c1b0aa7f6e399ec0f76119dc87/doc/doc_ch/PPOCRv3_det_train.md)
里面提到训练教师模型，一开始只让下载ResNet50_vd的预训练模型，没要求下载MobileNetV3的预…
-
Hi,
As the onnxruntime cannot take `torch.Tensor` as input, we should add a type check before we feed the `dataloader`'s output to onnxruntime inference session. I also found another bug when assigni…
-
pp-OCRv3训练时默认
Backbone:
name: MobileNetV3
scale: 0.5
model_name: large
如果更改为其他的scale例如1.0，下载预训练模型MobileNetV3_large_x1_0_pretrained参数会出现不匹配的情况。
只有默认的预训练模型参数才能匹配上，这种情况如何解决？
-
请提供下述完整信息以便快速定位问题/Please provide the following information to quickly locate the problem
- 系统环境/System Environment：windows10，python3.7
- 版本号/Version：Paddle：2.6 PaddleOCR：2.6 问题相关组件/Related compon…
-
I'm training a model that recognizes sentences. I have 500.000 senteneces.
This is the configuration of my model.
The model only predicts the **first word of a sentence.**
Does any one knows t…
-
Hello,
I first converted the mobilenetv3 large tensorflow model with post-training quantization to full integer according to https://www.tensorflow.org/lite/performance/post_training_quantization.…
-
### Checklist
- [X] I have searched related issues but cannot get the expected help.
- [X] 2. I have read the [FAQ documentation](https://github.com/open-mmlab/mmdeploy/tree/main/docs/en/faq.md) but …
-
I am so confused when I have a summary of MobileNetv3
The result is
![image](https://…