Update STPoseNet/README.md

STPoseNet/README.md

@@ -19,13 +19,11 @@ Keypoints identification program for STposeNet
 ```
 ### **Dataset preparation** 
 
-In this folder [Neurofinder](https://github.com/XZH-James/NeuroSeg2/tree/main/NeuroSeg%E2%85%A1-main/NeuroSeg%E2%85%A1-main/Neurofinder), We provide two images for testing.  
-* In this folder, [leftImg8bit](https://github.com/XZH-James/NeuroSeg2/tree/main/NeuroSeg%E2%85%A1-main/NeuroSeg%E2%85%A1-main/Neurofinder/test/leftImg8bit) stores the two-photon calcium imaging and [gtFine](https://github.com/XZH-James/NeuroSeg2/tree/main/NeuroSeg%E2%85%A1-main/NeuroSeg%E2%85%A1-main/Neurofinder/test/gtFine) stores the corresponding GT.  
-* [generate_dataset.py](https://github.com/XZH-James/NeuroSeg2/blob/main/NeuroSeg%E2%85%A1-main/NeuroSeg%E2%85%A1-main/Neurofinder/generate_dataset.py) is used to generate [image list](https://github.com/XZH-James/NeuroSeg2/tree/main/NeuroSeg%E2%85%A1-main/NeuroSeg%E2%85%A1-main/Neurofinder/imglists). After adding new images, run this code to generate the list for training and test code can read new images.  
+In this folder [datasets](https://huggingface.co/lvrgb777/STPoseNet/tree/main/dataset), We provide train dataset and test images and video.  
 
-### **Model preparation** 
+### **Pretrain weight preparation** 
 
-This folder [models](https://github.com/XZH-James/NeuroSeg2/tree/main/NeuroSeg%E2%85%A1-main/NeuroSeg%E2%85%A1-main/models) is used to store the pretrained model and the test model. The test model can be downloaded from our [huggingface](https://huggingface.co/XZH-James/NeuroSeg2/tree/main).  
+This file [weight](https://huggingface.co/lvrgb777/STPoseNet/blob/main/yolov8l_pose_mouse_com.pt) ,We provide pretrain weight
 
 ### **training or testing** 
 
@@ -33,23 +31,14 @@ After abtaining the dataset and model, running [test.py](https://github.com/XZH-
 Running [train.py](https://github.com/XZH-James/NeuroSeg2/blob/main/NeuroSeg%E2%85%A1-main/NeuroSeg%E2%85%A1-main/train.py) to train the new dataset.
 
 ### **The result** 
-
-[logs/evalution](https://github.com/XZH-James/NeuroSeg2/tree/main/NeuroSeg%E2%85%A1-main/NeuroSeg%E2%85%A1-main/logs/evalution/Neurofinder) contains the results of the neurons segmentation of NeuroSeg-II.  
-* [plt/difference](https://github.com/XZH-James/NeuroSeg2/tree/main/NeuroSeg%E2%85%A1-main/NeuroSeg%E2%85%A1-main/logs/evalution/Neurofinder/plt/difference) stores the segmented image by NeuroSeg-II.  
-* [evaluation_log.csv](https://github.com/XZH-James/NeuroSeg2/blob/main/NeuroSeg%E2%85%A1-main/NeuroSeg%E2%85%A1-main/logs/evalution/Neurofinder/evaluation_log.csv) is the score for this test.  
+2. run ./tool/min_img_label.py to perform data enhancement
+3. run mouse-train.py to train weight for your experimental data
+4. run mouse-pre to forecast result
 
 ## Other matters
-
-### **Core code** 
-In [neuroseg2](https://github.com/XZH-James/NeuroSeg2/tree/main/NeuroSeg%E2%85%A1-main/NeuroSeg%E2%85%A1-main/neuroseg2) are the core code of NeuroSeg-II.  
-* [model.py](https://github.com/XZH-James/NeuroSeg2/blob/main/NeuroSeg%E2%85%A1-main/NeuroSeg%E2%85%A1-main/neuroseg2/model.py) and [utils.py](https://github.com/XZH-James/NeuroSeg2/blob/main/NeuroSeg%E2%85%A1-main/NeuroSeg%E2%85%A1-main/neuroseg2/utils.py) are the code of overall structure.  
-* [Down.py](https://github.com/XZH-James/NeuroSeg2/blob/main/NeuroSeg%E2%85%A1-main/NeuroSeg%E2%85%A1-main/neuroseg2/Down.py) is the code of FPN+.  
-* [attention.py](https://github.com/XZH-James/NeuroSeg2/blob/main/NeuroSeg%E2%85%A1-main/NeuroSeg%E2%85%A1-main/neuroseg2/attention.py) is the code of attention mechanism.  
-* [visualize.py](https://github.com/XZH-James/NeuroSeg2/blob/main/NeuroSeg%E2%85%A1-main/NeuroSeg%E2%85%A1-main/neuroseg2/visualize.py) is the code for visual segmentation result.  
-
-### **Code of preprocessing** 
-In [utilities](https://github.com/XZH-James/NeuroSeg2/tree/main/NeuroSeg%E2%85%A1-main/NeuroSeg%E2%85%A1-main/utilities) are the code for preprocessing.  
+## Core module code location
+The main modifications base on yolo v8  are located ./ultralytice/engine/predictor
 
 ## Contact information
 
-If you have any questions about this project, please feel free to contact us. Email address: zhehao_xu@qq.com
+If you have any questions about this project, please feel free to contact us. Email address: 2245162223@qq.com