diff --git a/configs/picodet/README.md b/configs/picodet/README.md index 235b494e03845af9582f80cb55f51c34ebf5a9e7..cd25cfe64f0beec341049f8b8038919e707636b6 100644 --- a/configs/picodet/README.md +++ b/configs/picodet/README.md @@ -3,19 +3,19 @@ ![](../../docs/images/picedet_demo.jpeg) ## Introduction -We developed a series of lightweight models, which named `PP-PicoDet`. Because of its excellent performance, it is very suitable for deployment on mobile or CPU. For more details, please refer to our [report on arXiv](https://arxiv.org/abs/2111.00902). +We developed a series of lightweight models, named `PP-PicoDet`. Because of the excellent performance, our models are very suitable for deployment on mobile or CPU. For more details, please refer to our [report on arXiv](https://arxiv.org/abs/2111.00902). - 🌟 Higher mAP: the **first** object detectors that surpass mAP(0.5:0.95) **30+** within 1M parameters when the input size is 416. - 🚀 Faster latency: 150FPS on mobile ARM CPU. - 😊 Deploy friendly: support PaddleLite/MNN/NCNN/OpenVINO and provide C++/Python/Android implementation. -- 😍 Advanced algorithm: use the most advanced algorithms and innovate, such as ESNet, CSP-PAN, SimOTA with VFL, etc. +- 😍 Advanced algorithm: use the most advanced algorithms and offer innovation, such as ESNet, CSP-PAN, SimOTA with VFL, etc.
-### Comming soon +### Comming Soon - [ ] More series of model, such as smaller or larger model. - [ ] Pretrained models for more scenarios. - [ ] More features in need. @@ -35,7 +35,7 @@ We developed a series of lightweight models, which named `PP-PicoDet`. Because o | PicoDet-L | 416*416 | 36.6 | 52.5 | 3.30 | 3.76 | 23.36 | **21.85** | [model](https://paddledet.bj.bcebos.com/models/picodet_l_416_coco.pdparams) | [log](https://paddledet.bj.bcebos.com/logs/train_picodet_l_416_coco.log) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.3/configs/picodet/picodet_l_416_coco.yml) | | PicoDet-L | 640*640 | 40.9 | 57.6 | 3.30 | 8.91 | 54.11 | **50.55** | [model](https://paddledet.bj.bcebos.com/models/picodet_l_640_coco.pdparams) | [log](https://paddledet.bj.bcebos.com/logs/train_picodet_l_640_coco.log) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.3/configs/picodet/picodet_l_640_coco.yml) | -#### More config +#### More Configs | Model | Input size | mAPval
0.5:0.95 | mAPval
0.5 | Params
(M) | FLOPS
(G) | Latency[NCNN](#latency)
(ms) | Latency[Lite](#latency)
(ms) | download | config | | :--------------------------- | :--------: | :---------------------: | :----------------: | :----------------: | :---------------: | :-----------------------------: | :-----------------------------: | :-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------: | :------------------------------------------------------------------------------------------------------------------------------------------- | @@ -52,10 +52,26 @@ We developed a series of lightweight models, which named `PP-PicoDet`. Because o +#### Benchmark of Other Models + +| Model | Input size | mAPval
0.5:0.95 | mAPval
0.5 | Params
(M) | FLOPS
(G) | Latency[NCNN](#latency)
(ms) | +| :-------- | :--------: | :---------------------: | :----------------: | :----------------: | :---------------: | :-----------------------------: | +| YOLOv3-Tiny | 416*416 | 16.6 | 33.1 | 8.86 | 5.62 | 25.42 | +| YOLOv4-Tiny | 416*416 | 21.7 | 40.2 | 6.06 | 6.96 | 23.69 | +| PP-YOLO-Tiny | 320*320 | 20.6 | - | 1.08 | 0.58 | 6.75 | +| PP-YOLO-Tiny | 416*416 | 22.7 | - | 1.08 | 1.02 | 10.48 | +| Nanodet-M | 320*320 | 20.6 | - | 0.95 | 0.72 | 8.71 | +| Nanodet-M | 416*416 | 23.5 | - | 0.95 | 1.2 | 13.35 | +| Nanodet-M 1.5x | 416*416 | 26.8 | - | 2.08 | 2.42 | 15.83 | +| YOLOX-Nano | 416*416 | 25.8 | - | 0.91 | 1.08 | 19.23 | +| YOLOX-Tiny | 416*416 | 32.8 | - | 5.06 | 6.45 | 32.77 | +| YOLOv5n | 640*640 | 28.4 | 46.0 | 1.9 | 4.5 | 40.35 | +| YOLOv5s | 640*640 | 37.2 | 56.0 | 7.2 | 16.5 | 78.05 | + ## Deployment -### Export and Convert model +### Export and Convert Model
1. Export model (click to expand) @@ -131,14 +147,13 @@ paddle2onnx --model_dir output_inference/picodet_s_320_coco/ \ - [Android demo](https://github.com/JiweiMaster/PP-PicoDet-Android-Demo) +Android demo visualization:
-## Slim - -### quantization +## Quantization
Requirements: @@ -176,11 +191,11 @@ python tools/post_quant.py -c configs/picodet/picodet_s_320_coco.yml \ --slim_config configs/slim/post_quant/picodet_s_ptq.yml ``` -- Notes: Now the accuracy of post quant is abnormal and it is being debugged. +- Notes: Now the accuracy of post quant is abnormal and this problem is being solved.
-## Cite PiocDet +## Cite PP-PiocDet If you use PiocDet in your research, please cite our work by using the following BibTeX entry: ``` @misc{yu2021pppicodet, diff --git a/deploy/lite/README.md b/deploy/lite/README.md index dd3ddc5c91a99b22723f51163e16c2a4b972dea1..1b2adac6ee270bfade6e69a1d84f408ff8cd125b 100644 --- a/deploy/lite/README.md +++ b/deploy/lite/README.md @@ -22,9 +22,11 @@ Paddle Lite是飞桨轻量化推理引擎，为手机、IOT端提供高效推理 预测库有两种获取方式： 1. [**建议**]直接下载，预测库下载链接如下： - |平台|预测库下载链接| - |-|-| - |Android|[arm7](https://github.com/PaddlePaddle/Paddle-Lite/releases/download/v2.10-rc/inference_lite_lib.android.armv7.clang.c++_static.with_extra.with_cv.tar.gz) / [arm8](https://github.com/PaddlePaddle/Paddle-Lite/releases/download/v2.10-rc/inference_lite_lib.android.armv8.clang.c++_static.with_extra.with_cv.tar.gz)| + |平台| 架构 | 预测库下载链接| + |-|-|-| + |Android| arm7 | [inference_lite_lib](https://github.com/PaddlePaddle/Paddle-Lite/releases/download/v2.10-rc/inference_lite_lib.android.armv7.clang.c++_static.with_extra.with_cv.tar.gz) | + | Android | arm8 | [inference_lite_lib](https://github.com/PaddlePaddle/Paddle-Lite/releases/download/v2.10-rc/inference_lite_lib.android.armv8.clang.c++_static.with_extra.with_cv.tar.gz) | + | Android | arm8(FP16) | [inference_lite_lib](https://github.com/PaddlePaddle/Paddle-Lite/releases/download/v2.10-rc/inference_lite_lib.android.armv8_clang_c++_static_with_extra_with_cv_with_fp16.tiny_publish.zip) | **注意**：1. 如果是从 Paddle-Lite [官方文档](https://paddle-lite.readthedocs.io/zh/latest/quick_start/release_lib.html#android-toolchain-gcc)下载的预测库，注意选择`with_extra=ON，with_cv=ON`的下载链接。2. 目前只提供Android端demo，IOS端demo可以参考[Paddle-Lite IOS demo](https://github.com/PaddlePaddle/Paddle-Lite-Demo/tree/master/PaddleLite-ios-demo) @@ -35,7 +37,10 @@ git clone https://github.com/PaddlePaddle/Paddle-Lite.git cd Paddle-Lite # 如果使用编译方式，建议使用develop分支编译预测库 git checkout develop -./lite/tools/build_android.sh --arch=armv8 --with_cv=ON --with_extra=ON +# FP32 +./lite/tools/build_android.sh --arch=armv8 --toolchain=clang --with_cv=ON --with_extra=ON +# FP16 +./lite/tools/build_android.sh --arch=armv8 --toolchain=clang --with_cv=ON --with_extra=ON --with_arm82_fp16=ON ``` **注意**：编译Paddle-Lite获得预测库时，需要打开`--with_cv=ON --with_extra=ON`两个选项，`--arch`表示`arm`版本，这里指定为armv8，更多编译命令介绍请参考[链接](https://paddle-lite.readthedocs.io/zh/latest/source_compile/compile_andriod.html#id2)。 @@ -131,7 +136,10 @@ python tools/export_model.py -c configs/picodet/picodet_s_320_coco.yml \ -o weights=https://paddledet.bj.bcebos.com/models/picodet_s_320_coco.pdparams --output_dir=output_inference # 将inference模型转化为Paddle-Lite优化模型 +# FP32 paddle_lite_opt --valid_targets=arm --model_file=output_inference/picodet_s_320_coco/model.pdmodel --param_file=output_inference/picodet_s_320_coco/model.pdiparams --optimize_out=output_inference/picodet_s_320_coco/model +# FP16 +paddle_lite_opt --valid_targets=arm --model_file=output_inference/picodet_s_320_coco/model.pdmodel --param_file=output_inference/picodet_s_320_coco/model.pdiparams --optimize_out=output_inference/picodet_s_320_coco/model --enable_fp16=true # 将inference模型配置转化为json格式 python deploy/lite/convert_yml_to_json.py output_inference/picodet_s_320_coco/infer_cfg.yml @@ -144,7 +152,7 @@ python deploy/lite/convert_yml_to_json.py output_inference/picodet_s_320_coco/in ### 2.2 与手机联调 首先需要进行一些准备工作。 -1. 准备一台arm8的安卓手机，如果编译的预测库和opt文件是armv7，则需要arm7的手机，并修改Makefile中`ARM_ABI = arm7`。 +1. 准备一台arm8的安卓手机，如果编译的预测库是armv7，则需要arm7的手机，并修改Makefile中`ARM_ABI=arm7`。 2. 电脑上安装ADB工具，用于调试。 ADB安装方式如下： 2.1. MAC电脑安装ADB: diff --git a/docs/images/picodet_android_demo1.jpg b/docs/images/picodet_android_demo1.jpg index 6f5d2bc6716c2b3edf2fc9bda74756fafc53283d..ecb5dae5135d1912ef65485e31e3b00715a8195b 100644 Binary files a/docs/images/picodet_android_demo1.jpg and b/docs/images/picodet_android_demo1.jpg differ diff --git a/docs/images/picodet_android_demo2.jpg b/docs/images/picodet_android_demo2.jpg index 3eb2d77f08d618a6160dcf74cb5447026aa8ad47..e08ffb8b0cb4859a76b62017c6beef951cbc2f85 100644 Binary files a/docs/images/picodet_android_demo2.jpg and b/docs/images/picodet_android_demo2.jpg differ