update readme

satchelwu · Sep 22, 2020 · bb64512 · bb64512
1 parent f88184e
commit bb64512
Show file tree

Hide file tree

Showing 8 changed files with 8 additions and 6 deletions.
diff --git a/README.md b/README.md
@@ -10,6 +10,7 @@ All the models are implemented in pytorch or mxnet first, and export a weights f
 
 ## News
 
+- `16 Sep 2020`. [hwh-hit](https://github.com/hwh-hit) added ufld(Ultra-Fast-Lane-Detection, ECCV2020).
 - `13 Sep 2020`. Add crnn, and got 1000fps on GTX1080.
 - `7 Sep 2020`. Implement retinaface(mobilenet0.25), and got 333fps on GTX1080.
 - `28 Aug 2020`. [BaofengZan](https://github.com/BaofengZan) added a tutorial for compiling and running tensorrtx on windows.
@@ -66,6 +67,7 @@ Following models are implemented.
 |[retinafaceAntiCov](./retinafaceAntiCov)| mobilenet0.25, weights from [deepinsight/insightface](https://github.com/deepinsight/insightface), retinaface anti-COVID-19, detect face and mask attribute |
 |[dbnet](./dbnet)| Scene Text Detection, weights from [BaofengZan/DBNet.pytorch](https://github.com/BaofengZan/DBNet.pytorch) |
 |[crnn](./crnn)| pytorch implementation from [meijieru/crnn.pytorch](https://github.com/meijieru/crnn.pytorch) |
+|[ufld](./ufld)| pytorch implementation from [Ultra-Fast-Lane-Detection](https://github.com/cfzd/Ultra-Fast-Lane-Detection), ECCV2020 |
 
 ## Tricky Operations
 

diff --git a/lane_det/CMakeLists.txt → ufld/CMakeLists.txt b/lane_det/CMakeLists.txt → ufld/CMakeLists.txt
diff --git a/lane_det/README.md → ufld/README.md b/lane_det/README.md → ufld/README.md
@@ -1,4 +1,4 @@
-# Ultra-Fast-Lane-Detection
+# Ultra-Fast-Lane-Detection(UFLD)
 
 The Pytorch implementation is [Ultra-Fast-Lane-Detection](https://github.com/cfzd/Ultra-Fast-Lane-Detection).
 
@@ -9,7 +9,7 @@ The Pytorch implementation is [Ultra-Fast-Lane-Detection](https://github.com/cfz
 git clone https://github.com/wang-xinyu/tensorrtx.git
 git clone https://github.com/cfzd/Ultra-Fast-Lane-Detection.git
 // download its weights 'tusimple_18.pth'
-// copy tensorrtx/lane_det/gen_wts.py into Ultra-Fast-Lane-Detection/
+// copy tensorrtx/ufld/gen_wts.py into Ultra-Fast-Lane-Detection/
 // ensure the file name is tusimple_18.pth and lane.wts in gen_wts.py
 // go to Ultra-Fast-Lane-Detection
 python gen_wts.py
@@ -18,18 +18,18 @@ python gen_wts.py
 python pth2onnx.py
 //a file 'lane.onnx' will be generated.
 
-2. build tensorrtx/lane_det and run
+2. build tensorrtx/ufld and run
 
 mkdir build
 cd build
 cmake ..
 make
-sudo ./lane_det -s             // serialize model to plan file i.e. 'lane.engine'
-sudo ./lane_det -d  ../data // deserialize plan file and run inference, the images in data will be processed.
+sudo ./lane_det -s          // serialize model to plan file i.e. 'lane.engine'
+sudo ./lane_det -d  PATH_TO_YOUR_IMAGE_FOLDER // deserialize plan file and run inference, the images will be processed.
 
 ```
 
 ## More Information
-1. Changed the preprocess and postprocess in tensorrtx, give a different way to convert NHWC to NCHW in preprocess and just show the reslut using opencv rather than saving the result in postprocess.
+1. Changed the preprocess and postprocess in tensorrtx, give a different way to convert NHWC to NCHW in preprocess and just show the result using opencv rather than saving the result in postprocess.
 2. If there are some bugs where you inference with multi batch_size, just modify the code in preprocess or postprocess, it's not complicated.
 3. Some results are stored in resluts folder.
diff --git a/lane_det/common.hpp → ufld/common.hpp b/lane_det/common.hpp → ufld/common.hpp
diff --git a/lane_det/gen_wts.py → ufld/gen_wts.py b/lane_det/gen_wts.py → ufld/gen_wts.py
diff --git a/lane_det/lane_det.cpp → ufld/lane_det.cpp b/lane_det/lane_det.cpp → ufld/lane_det.cpp
diff --git a/lane_det/logging.h → ufld/logging.h b/lane_det/logging.h → ufld/logging.h
diff --git a/lane_det/pth2onnx.py → ufld/pth2onnx.py b/lane_det/pth2onnx.py → ufld/pth2onnx.py