mirror of
https://github.com/PaddlePaddle/FastDeploy.git
synced 2025-10-06 00:57:33 +08:00
820a5c5647
* Update README.md * Update README.md * Update README.md * Create README.md * Update README.md * Update README.md * Update README.md * Update README.md * Add evaluation calculate time and fix some bugs * Update classification __init__ * Move to ppseg * Add segmentation doc * Add PaddleClas infer.py * Update PaddleClas infer.py * Delete .infer.py.swp * Add PaddleClas infer.cc * Update README.md * Update README.md * Update README.md * Update infer.py * Update README.md * Update README.md * Update README.md * Update README.md * Update README.md * Update README.md * Update README.md * Update README.md * Update README.md * Update README.md * Update README.md * Update README.md * Update README.md * Update README.md * Add PaddleSeg doc and infer.cc demo * Update README.md * Update README.md * Update README.md * Update README.md * Update README.md * Update README.md * Create segmentation_result.md * Update README.md * Update segmentation_result.md * Update segmentation_result.md * Update segmentation_result.md * Update classification and detection evaluation function * Fix python grammar bug * Add PPSeg evaluation function * Add average_inference_time function * Update PaddleSeg example infer.py Tensorrt setting * Fix bug for CI convert os.system to subprocess.popen * Fix tensorrt cann't link to *.so * Fix tensorrt can't find so bug * Update CMakeLists * Update CMakeLists * Update CMakeLists * Update CMakeLists * Update README.md * Update README.md * Update README.md * Update setup.py Add --force-rpath Co-authored-by: Jason <jiangjiajun@baidu.com> Co-authored-by: felixhjh <852142024@example.com>
PaddleSeg 模型部署
模型版本说明
目前FastDeploy支持如下模型的部署
准备PaddleSeg部署模型
PaddleSeg模型导出，请参考其文档说明模型导出
注意：在使用PaddleSeg模型导出时，可指定--input_shape参数，若预测输入图片尺寸并不固定，建议使用默认值即不指定该参数。PaddleSeg导出的模型包含model.pdmodel、model.pdiparams和deploy.yaml三个文件，FastDeploy会从yaml文件中获取模型在推理时需要的预处理信息。
下载预训练模型
为了方便开发者的测试，下面提供了PaddleSeg导出的部分模型（导出方式为：不指定input_shape和with_softmax，指定without_argmax），开发者可直接下载使用。
| 模型 | 参数文件大小 | 输入Shape | mIoU | mIoU (flip) | mIoU (ms+flip) |
|---|---|---|---|---|---|
| Unet-cityscapes | 52MB | 1024x512 | 65.00% | 66.02% | 66.89% |
| PP-LiteSeg-T(STDC1)-cityscapes | 31MB | 1024x512 | 73.10% | 73.89% | - |
| PP-HumanSegV1-Lite | 543KB | 192x192 | 86.2% | - | - |
| PP-HumanSegV1-Server | 103MB | 512x512 | 96.47% | - | - |
| FCN-HRNet-W18-cityscapes | 37MB | 1024x512 | 78.97% | 79.49% | 79.74% |
| Deeplabv3-ResNet50-OS8-cityscapes | 150MB | 1024x512 | 79.90% | 80.22% | 80.47% |