@@ -244,6 +244,7 @@
- `Cascade-Faster-RCNN`为`Cascade-Faster-RCNN-ResNet50vd-DCN`，PaddleDetection将其优化到COCO数据mAP为47.8%时推理速度为20FPS
- `PP-YOLO`在COCO数据集精度45.9%，Tesla V100预测速度72.9FPS，精度速度均优于[YOLOv4](https://arxiv.org/abs/2004.10934)
- `PP-YOLO v2`是对`PP-YOLO`模型的进一步优化，在COCO数据集精度49.5%，Tesla V100预测速度68.9FPS
+- `PP-YOLOE`是对`PP-YOLO v2`模型的进一步优化，在COCO数据集精度51.4%，Tesla V100预测速度78.1FPS
- 图中模型均可在[模型库](#模型库)中获取
各移动端模型在COCO数据集上精度mAP和高通骁龙865处理器上预测速度(FPS)对比图。
@@ -269,16 +270,16 @@
### 进阶教程
- 参数配置
-
+
- [RCNN参数说明](docs/tutorials/config_annotation/faster_rcnn_r50_fpn_1x_coco_annotation.md)
- [PP-YOLO参数说明](docs/tutorials/config_annotation/ppyolo_r50vd_dcn_1x_coco_annotation.md)
- 模型压缩(基于[PaddleSlim](https://github.com/PaddlePaddle/PaddleSlim))
-
+
- [剪裁/量化/蒸馏教程](configs/slim)
- [推理部署](deploy/README.md)
-
+
- [模型导出教程](deploy/EXPORT_MODEL.md)
- [Paddle Inference部署](deploy/README.md)
- [Python端推理部署](deploy/python)
@@ -289,7 +290,7 @@
- [推理benchmark](deploy/BENCHMARK_INFER.md)
- 进阶开发
-
+
- [数据处理模块](docs/advanced_tutorials/READER.md)
- [新增检测模型](docs/advanced_tutorials/MODEL_TECHNICAL.md)
diff --git a/README_en.md b/README_en.md
index ade0b769bfe96436b2bd2ce4a93c569029d54f9f..c9e792199878c180176a442689e6006dbcc08d92 100644
--- a/README_en.md
+++ b/README_en.md
@@ -17,7 +17,7 @@ English | [简体中文](README_cn.md)
##
Latest News
- 🔥 **2022.3.24：PaddleDetection [release 2.4 version](https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.4)**
-
+
- Release GPU SOTA object detection series models (s/m/l/x) [PP-YOLOE](configs/ppyoloe), achieving mAP as 51.4% on COCO test dataset and 78.1 FPS on Nvidia V100, supporting AMP training and its training speed is 33% faster than PP-YOLOv2.
- Release enhanced models of [PP-PicoDet](configs/picodet), including PP-PicoDet-XS model with 0.7M parameters, its mAP promoted ~2% on COCO, inference speed accelerated 63% on CPU, and post-processing integrated into the network to optimize deployment pipeline.
- Release real-time human analysis tool [PP-Human](deploy/pphuman), which is based on data from real-life situations, supporting pedestrian detection, attribute recognition, human tracking, multi-camera tracking, human statistics and action recognition.
@@ -46,20 +46,20 @@ PaddleDetection is an end-to-end object detection development kit based on Paddl
##
Features
-- **Rich Models**
-
+- **Rich Models**
+
PaddleDetection provides rich of models, including **250+ pre-trained models** such as **object detection**, **instance segmentation**, **face detection**, **keypoint detection**, **multi-object tracking** and etc, covering a variety of **global competition champion** schemes.
- **Highly Flexible**
-
+
Components are designed to be modular. Model architectures, as well as data preprocess pipelines and optimization strategies, can be easily customized with simple configuration changes.
-- **Production Ready**
-
+- **Production Ready**
+
From data augmentation, constructing models, training, compression, depolyment, get through end to end, and complete support for multi-architecture, multi-device deployment for **cloud and edge device**.
-- **High Performance**
-
+- **High Performance**
+
Based on the high performance core of PaddlePaddle, advantages of training speed and memory occupation are obvious. FP16 training and multi-machine training are supported as well.
##
Community
@@ -67,7 +67,7 @@ PaddleDetection is an end-to-end object detection development kit based on Paddl
- If you have any problem or suggestion on PaddleDetection, please send us issues through [GitHub Issues](https://github.com/PaddlePaddle/PaddleDetection/issues).
- Welcome to Join PaddleDetection QQ Group and Wechat Group (reply "Det").
-
+
@@ -243,7 +243,7 @@ The relationship between COCO mAP and FPS on Tesla V100 of representative models
- `PP-YOLO` achieves mAP of 45.9% on COCO and 72.9FPS on Tesla V100. Both precision and speed surpass [YOLOv4](https://arxiv.org/abs/2004.10934)
- `PP-YOLO v2` is optimized version of `PP-YOLO` which has mAP of 49.5% and 68.9FPS on Tesla V100
-
+- `PP-YOLOE` is optimized version of `PP-YOLO v2` which has mAP of 51.4% and 78.1FPS on Tesla V100
- All these models can be get in [Model Zoo](#ModelZoo)
The relationship between COCO mAP and FPS on Qualcomm Snapdragon 865 of representative mobile side models.
@@ -268,16 +268,16 @@ The relationship between COCO mAP and FPS on Qualcomm Snapdragon 865 of represen
### Advanced Tutorials
- Parameter configuration
-
+
- [Parameter configuration for RCNN model](docs/tutorials/config_annotation/faster_rcnn_r50_fpn_1x_coco_annotation_en.md)
- [Parameter configuration for PP-YOLO model](docs/tutorials/config_annotation/ppyolo_r50vd_dcn_1x_coco_annotation_en.md)
- Model Compression(Based on [PaddleSlim](https://github.com/PaddlePaddle/PaddleSlim))
-
+
- [Prune/Quant/Distill](configs/slim)
- Inference and deployment
-
+
- [Export model for inference](deploy/EXPORT_MODEL_en.md)
- [Paddle Inference](deploy/README_en.md)
- [Python inference](deploy/python)
@@ -289,7 +289,7 @@ The relationship between COCO mAP and FPS on Qualcomm Snapdragon 865 of represen
- [Exporting to ONNX and using OpenVINO for inference](docs/advanced_tutorials/openvino_inference/README.md)
- Advanced development
-
+
- [New data augmentations](docs/advanced_tutorials/READER_en.md)
- [New detection algorithms](docs/advanced_tutorials/MODEL_TECHNICAL.md)
diff --git a/docs/images/fps_map.png b/docs/images/fps_map.png
index d73877729c0775709e5954c008a88776bf48606a..44f3e846a877fd08fcd905027a1eced14ccb5539 100644
Binary files a/docs/images/fps_map.png and b/docs/images/fps_map.png differ