diff --git a/README.md b/README.md
index 0a019b54..f4ee8439 100644
--- a/README.md
+++ b/README.md
@@ -47,7 +47,7 @@
 
                 2. 填写`jd_wstool` 监听地址ip
 
-                   如：监听地址1： http://192.168.0.101:5201，填在脚本开头 ipaddr= '192.168.0.101'
+                   如：监听地址1： http://192.168.0.101:5201 ，填在脚本开头 ipaddr= '192.168.0.101'
 
                 3. 运行脚本
 
@@ -97,9 +97,13 @@
 
 3. 第二关：图形验证码（任选以下一种类型，默认采用本地识别）
 
-    1. 本地识别（再也不用花钱了👍），来自[@AntonVanke](https://github.com/AntonVanke)
+    1. 本地识别（再也不用花钱了👍）
+       
+        * 来自[@AntonVanke](https://github.com/AntonVanke)
        大佬提供的 [JDCaptcha](https://github.com/AntonVanke/JDCaptcha) 项目(已集成)
-       ，[测试图在最后一张](https://github.com/yqchilde/JDMemberCloseAccount#screenshots) ，
+       ，[测试图在最后一张](https://github.com/yqchilde/JDMemberCloseAccount#screenshots) 
+          
+        * [@dd178](https://github.com/dd178) 使用 [yolov4](https://github.com/AlexeyAB/darknet) 训练的权重
 
     2. 收费的打码平台
 
@@ -131,12 +135,19 @@
    pip3 install -r requirements.txt -i https://mirrors.aliyun.com/pypi/simple/
    ```
 
-3. 下载对应的浏览器驱动放到项目的`drivers`文件夹下面
+3. 下载必要文件
 
-    * `chrome`请访问`chrome://version/`查看浏览器的版本，然后去 [chromedriver](http://chromedriver.storage.googleapis.com/index.html)
+    3.1 下载对应的浏览器驱动放到项目的`drivers`文件夹下面
+      
+    `chrome`请访问`chrome://version/`查看浏览器的版本，然后去 [chromedriver](http://chromedriver.storage.googleapis.com/index.html)
       下载对应的版本/系统驱动（只需要保证版本号前三段一致即可，比如`91.0.4472.77`只需要保证`91.0.4472.x`就行），下载后解压，将其可执行文件（mac为`chromedriver`
       ，win为`chromedriver.exe`放在项目的`drivers`目录下即可）
+   
+    3.2 如果想使用yolov4识别验证码
 
+    下载[权重文件](https://github.com/dd178/JDMemberCloseAccount/releases/download/v1.0.3/yolov4-custom.tar.gz) ，将`yolov4-custom.weights`解压至`yolov4`文件夹下
+   
+    
 ### 2. 补充配置文件
 
 * `config.yaml`文件
@@ -193,7 +204,7 @@ sms_captcha:
     aliyun_appcode: ""
 
 # image_captcha 图形验证码相关
-# image_captcha.type: 图形验证码类型，可选：local、cjy、tj
+# image_captcha.type: 图形验证码类型，可选：local、cjy、tj、yolov4
 # image_captcha.cjy_username: 超级鹰账号，仅在 image_captcha.type 为 cjy 时需要设置
 # image_captcha.cjy_password: 超级鹰密码，仅在 image_captcha.type 为 cjy 时需要设置
 # image_captcha.cjy_soft_id: 超级鹰软件ID，仅在 image_captcha.type 为 cjy 时需要设置
@@ -201,6 +212,9 @@ sms_captcha:
 # image_captcha.tj_username: 图鉴账号，仅在 image_captcha.type 为 tj 时需要设置
 # image_captcha.tj_password: 图鉴密码，仅在 image_captcha.type 为 tj 时需要设置
 # image_captcha.tj_type_id: 图鉴验证码类型，仅在 image_captcha.type 为 tj 时需要设置，且该项目指定为 19
+# yolov4_weights: yolov4权重文件路径，仅在 image_captcha.type 为 yolov4 时需要设置
+# yolov4_cfg: yolov4配置文件路径，仅在 image_captcha.type 为 yolov4 时需要设置
+# CUDA: 尝试使用CUDA加速，据说速度可提升几倍到几十倍，需要编译安装opencv，仅在 image_captcha.type 为 yolov4 时需要设置
 image_captcha:
   type: "local"
   cjy_username: ""
@@ -210,6 +224,9 @@ image_captcha:
   tj_username: ""
   tj_password: ""
   tj_type_id: 19
+  yolov4_weights: "yolov4/yolov4-custom.weights"
+  yolov4_cfg: "yolov4/yolov4-custom.cfg"
+  CUDA: false
 
 # user-agent 用户代理，可自行配置
 user-agent:
diff --git a/captcha/jd_yolo_captcha.py b/captcha/jd_yolo_captcha.py
new file mode 100644
index 00000000..6ab558bf
--- /dev/null
+++ b/captcha/jd_yolo_captcha.py
@@ -0,0 +1,72 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+# @Time    : 2021-06-21 11:33
+# @Author  : 178
+import cv2
+import base64
+import numpy as np
+import os
+import sys
+
+class JDyolocaptcha(object):
+    """
+    yolov4类
+    """
+    def __init__(self, _config):
+        self.CONFIDENCE_THRESHOLD = 0.8  # 最低置信度
+        self.NMS_THRESHOLD = 0.01  # 去除重复匹配
+        from utils.logger import Log
+        self.logger = Log().logger
+        weights = _config['yolov4_weights']
+        cfg = _config['yolov4_cfg']
+        if os.path.exists(weights):
+            self.net = cv2.dnn.readNet(weights, cfg)
+        else:
+            self.logger.error("找不到权重文件")
+            sys.exit(1)
+        if _config['CUDA']:
+            self.net.setPreferableBackend(cv2.dnn.DNN_BACKEND_CUDA)
+            self.net.setPreferableTarget(cv2.dnn.DNN_TARGET_CUDA)
+        # 由于两张图片大小不一样，因此要使用两个不同大小的网络去识别，否则识别率极低
+        self.cpc_model = cv2.dnn_DetectionModel(self.net)
+        self.pcp_model = cv2.dnn_DetectionModel(self.net)
+        self.cpc_model.setInputParams(size=(320, 320), scale=1/255, swapRB=True)  # size为32的倍数，越大越慢，但不一定识别率越高
+        self.pcp_model.setInputParams(size=(224, 128), scale=1/255, swapRB=True)  # size为32的倍数
+
+
+    def base64_conversion(self, data):
+        """
+        base64转Mat
+        :param data:
+        :return:
+        """
+        imgData = base64.b64decode(data.replace("data:image/jpg;base64,", ""))
+        nparr = np.frombuffer(imgData, np.uint8)
+        return cv2.imdecode(nparr, cv2.IMREAD_COLOR)
+
+
+    def identify(self, cpc, pcp):
+        """
+        识别验证码并返回坐标
+        :param cpc:
+        :param pcp:
+        :return:
+        """
+        try:
+            cpc_classes, cpc_scores, cpc_boxes = self.cpc_model.detect(cpc, self.CONFIDENCE_THRESHOLD, self.NMS_THRESHOLD)
+            pcp_classes, pcp_scores, pcp_boxes = self.pcp_model.detect(pcp, self.CONFIDENCE_THRESHOLD, self.NMS_THRESHOLD)
+            if pcp_classes[0] in cpc_classes:  # 判断识别小图的结果是否在大图里面
+                x1, y1, x2, y2 = cpc_boxes[cpc_classes.tolist().index(pcp_classes[0])]
+                if x2 - x1 < 200:  # 防止结果为背景，因此要剔除x差值在200以上的结果
+                    r = (x1*2+x2)//2, (y1*2+y2)//2
+                    return True, r
+                else:
+                    return False, (None, None)
+            else:
+                return False, (None, None)
+        except:
+            return False, (None, None)
+
+
+    def JDyolo(self, cpc_img_path_base64, pcp_show_picture_path_base64):
+        return self.identify(self.base64_conversion(cpc_img_path_base64), self.base64_conversion(pcp_show_picture_path_base64))
\ No newline at end of file
diff --git a/config.yaml b/config.yaml
index 2926e941..8f66d5b3 100644
--- a/config.yaml
+++ b/config.yaml
@@ -49,7 +49,7 @@ sms_captcha:
     aliyun_appcode: ""
 
 # image_captcha 图形验证码相关
-# image_captcha.type: 图形验证码类型，可选：local、cjy、tj
+# image_captcha.type: 图形验证码类型，可选：local、cjy、tj、yolov4
 # image_captcha.cjy_username: 超级鹰账号，仅在 image_captcha.type 为 cjy 时需要设置
 # image_captcha.cjy_password: 超级鹰密码，仅在 image_captcha.type 为 cjy 时需要设置
 # image_captcha.cjy_soft_id: 超级鹰软件ID，仅在 image_captcha.type 为 cjy 时需要设置
@@ -57,6 +57,9 @@ sms_captcha:
 # image_captcha.tj_username: 图鉴账号，仅在 image_captcha.type 为 tj 时需要设置
 # image_captcha.tj_password: 图鉴密码，仅在 image_captcha.type 为 tj 时需要设置
 # image_captcha.tj_type_id: 图鉴验证码类型，仅在 image_captcha.type 为 tj 时需要设置，且该项目指定为 19
+# yolov4_weights: yolov4权重文件路径，仅在 image_captcha.type 为 yolov4 时需要设置
+# yolov4_cfg: yolov4配置文件路径，仅在 image_captcha.type 为 yolov4 时需要设置
+# CUDA: 尝试使用CUDA加速，据说速度可提升几倍到几十倍，需要编译安装opencv，仅在 image_captcha.type 为 yolov4 时需要设置
 image_captcha:
   type: "local"
   cjy_username: ""
@@ -66,6 +69,9 @@ image_captcha:
   tj_username: ""
   tj_password: ""
   tj_type_id: 19
+  yolov4_weights: "yolov4/yolov4-custom.weights"
+  yolov4_cfg: "yolov4/yolov4-custom.cfg"
+  CUDA: false
 
 # user-agent 用户代理，可自行配置
 user-agent:
diff --git a/main.py b/main.py
index 28fc7942..6f66818b 100644
--- a/main.py
+++ b/main.py
@@ -10,6 +10,7 @@
 from captcha.chaojiying import ChaoJiYing
 from captcha.tujian import TuJian
 from captcha.jd_captcha import JDcaptcha_base64
+from captcha.jd_yolo_captcha import JDyolocaptcha
 from utils.logger import Log
 from utils.config import get_config
 from utils.selenium_browser import get_browser
@@ -93,6 +94,8 @@ def __init__(self):
             self.tj = TuJian(self.image_captcha_cfg)
         elif self.image_captcha_cfg["type"] == "local":
             pass
+        elif self.image_captcha_cfg["type"] == "yolov4":
+            self.JDyolo = JDyolocaptcha(self.image_captcha_cfg)
         else:
             WARN("请在config.yaml中补充image_captcha.type")
 
@@ -451,8 +454,12 @@ def local_auto_identify_captcha_click():
                             pcp_show_picture_path_base64 = self.wait.until(EC.presence_of_element_located(
                                 (By.XPATH, '//*[@class="pcp_showPicture"]'))).get_attribute('src')
                             # 正在识别验证码
-                            INFO("正在通过本地引擎识别")
-                            res = JDcaptcha_base64(cpc_img_path_base64, pcp_show_picture_path_base64)
+                            if self.image_captcha_cfg["type"] == "local":
+                                INFO("正在通过本地引擎识别")
+                                res = JDcaptcha_base64(cpc_img_path_base64, pcp_show_picture_path_base64)
+                            else:
+                                INFO("正在通过深度学习引擎识别")
+                                res = self.JDyolo.JDyolo(cpc_img_path_base64, pcp_show_picture_path_base64)
                             if res[0]:
                                 ActionChains(self.browser).move_to_element_with_offset(
                                     cpc_img, int(res[1][0] * zoom),
@@ -475,7 +482,7 @@ def local_auto_identify_captcha_click():
                         return False
 
                     # 识别点击，如果有一次失败将再次尝试一次，再失败就跳过
-                    if self.image_captcha_cfg["type"] == "local":
+                    if self.image_captcha_cfg["type"] in ["local", "yolov4"]:
                         if not local_auto_identify_captcha_click():
                             INFO("验证码位置点击错误，尝试再试一次")
                             local_auto_identify_captcha_click()
diff --git a/requirements.txt b/requirements.txt
index 7a71ae04..81799eb1 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -7,4 +7,5 @@ Pillow~=8.2.0
 numpy~=1.20.3
 urllib3~=1.26.5
 baidu-aip==2.2.18.0
-websockets~=9.1
\ No newline at end of file
+websockets~=9.1
+opencv_python~=4.5.2.54
\ No newline at end of file
diff --git a/yolov4/yolov4-custom.cfg b/yolov4/yolov4-custom.cfg
new file mode 100644
index 00000000..21fbfd3a
--- /dev/null
+++ b/yolov4/yolov4-custom.cfg
@@ -0,0 +1,1164 @@
+[net]
+# Testing
+#batch=1
+#subdivisions=1
+# Training
+batch=64
+subdivisions=32
+width=352
+height=352
+channels=3
+momentum=0.949
+decay=0.0005
+angle=0
+saturation = 1.5
+exposure = 1.5
+hue=.1
+flip=1
+#use_cuda_graph = 1
+
+# CUDNN_HALF
+loss_scale=128
+
+learning_rate=0.001
+burn_in=1000
+max_batches = 100000
+policy=steps
+steps=80000,90000
+scales=.1,.1
+
+#cutmix=1
+mosaic=1
+
+#:104x104 54:52x52 85:26x26 104:13x13 for 416
+
+[convolutional]
+batch_normalize=1
+filters=32
+size=3
+stride=1
+pad=1
+activation=mish
+
+# Downsample
+
+[convolutional]
+batch_normalize=1
+filters=64
+size=3
+stride=2
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=64
+size=1
+stride=1
+pad=1
+activation=mish
+
+[route]
+layers = -2
+
+[convolutional]
+batch_normalize=1
+filters=64
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=32
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=64
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=64
+size=1
+stride=1
+pad=1
+activation=mish
+
+[route]
+layers = -1,-7
+
+[convolutional]
+batch_normalize=1
+filters=64
+size=1
+stride=1
+pad=1
+activation=mish
+
+# Downsample
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=3
+stride=2
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=64
+size=1
+stride=1
+pad=1
+activation=mish
+
+[route]
+layers = -2
+
+[convolutional]
+batch_normalize=1
+filters=64
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=64
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=64
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=64
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=64
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=64
+size=1
+stride=1
+pad=1
+activation=mish
+
+[route]
+layers = -1,-10
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=mish
+
+# Downsample
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=3
+stride=2
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=mish
+
+[route]
+layers = -2
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=mish
+
+[route]
+layers = -1,-28
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=mish
+
+# Downsample
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=3
+stride=2
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=mish
+
+[route]
+layers = -2
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=mish
+
+[route]
+layers = -1,-28
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=1
+stride=1
+pad=1
+activation=mish
+
+# Downsample
+
+[convolutional]
+batch_normalize=1
+filters=1024
+size=3
+stride=2
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=1
+stride=1
+pad=1
+activation=mish
+
+[route]
+layers = -2
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=1
+stride=1
+pad=1
+activation=mish
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=3
+stride=1
+pad=1
+activation=mish
+
+[shortcut]
+from=-3
+activation=linear
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=1
+stride=1
+pad=1
+activation=mish
+
+[route]
+layers = -1,-16
+
+[convolutional]
+batch_normalize=1
+filters=1024
+size=1
+stride=1
+pad=1
+activation=mish
+stopbackward=800
+
+##########################
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=1
+pad=1
+filters=1024
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=1
+stride=1
+pad=1
+activation=leaky
+
+### SPP ###
+[maxpool]
+stride=1
+size=5
+
+[route]
+layers=-2
+
+[maxpool]
+stride=1
+size=9
+
+[route]
+layers=-4
+
+[maxpool]
+stride=1
+size=13
+
+[route]
+layers=-1,-3,-5,-6
+### End SPP ###
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=1
+pad=1
+filters=1024
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[upsample]
+stride=2
+
+[route]
+layers = 85
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[route]
+layers = -1, -3
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=1
+pad=1
+filters=512
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=1
+pad=1
+filters=512
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[upsample]
+stride=2
+
+[route]
+layers = 54
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[route]
+layers = -1, -3
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=1
+pad=1
+filters=256
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=1
+pad=1
+filters=256
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=128
+size=1
+stride=1
+pad=1
+activation=leaky
+
+##########################
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=1
+pad=1
+filters=256
+activation=leaky
+
+[convolutional]
+size=1
+stride=1
+pad=1
+filters=495
+activation=linear
+
+
+[yolo]
+mask = 0,1,2
+anchors = 12, 16, 19, 36, 40, 28, 36, 75, 76, 55, 72, 146, 142, 110, 192, 243, 459, 401
+classes=160
+num=9
+jitter=.3
+ignore_thresh = .7
+truth_thresh = 1
+scale_x_y = 1.2
+iou_thresh=0.213
+cls_normalizer=1.0
+iou_normalizer=0.07
+iou_loss=ciou
+nms_kind=greedynms
+beta_nms=0.6
+max_delta=5
+
+
+[route]
+layers = -4
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=2
+pad=1
+filters=256
+activation=leaky
+
+[route]
+layers = -1, -16
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=1
+pad=1
+filters=512
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=1
+pad=1
+filters=512
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=256
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=1
+pad=1
+filters=512
+activation=leaky
+
+[convolutional]
+size=1
+stride=1
+pad=1
+filters=495
+activation=linear
+
+
+[yolo]
+mask = 3,4,5
+anchors = 12, 16, 19, 36, 40, 28, 36, 75, 76, 55, 72, 146, 142, 110, 192, 243, 459, 401
+classes=160
+num=9
+jitter=.3
+ignore_thresh = .7
+truth_thresh = 1
+scale_x_y = 1.1
+iou_thresh=0.213
+cls_normalizer=1.0
+iou_normalizer=0.07
+iou_loss=ciou
+nms_kind=greedynms
+beta_nms=0.6
+max_delta=5
+
+
+[route]
+layers = -4
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=2
+pad=1
+filters=512
+activation=leaky
+
+[route]
+layers = -1, -37
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=1
+pad=1
+filters=1024
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=1
+pad=1
+filters=1024
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+filters=512
+size=1
+stride=1
+pad=1
+activation=leaky
+
+[convolutional]
+batch_normalize=1
+size=3
+stride=1
+pad=1
+filters=1024
+activation=leaky
+
+[convolutional]
+size=1
+stride=1
+pad=1
+filters=495
+activation=linear
+
+
+[yolo]
+mask = 6,7,8
+anchors = 12, 16, 19, 36, 40, 28, 36, 75, 76, 55, 72, 146, 142, 110, 192, 243, 459, 401
+classes=160
+num=9
+jitter=.3
+ignore_thresh = .7
+truth_thresh = 1
+random=1
+scale_x_y = 1.05
+iou_thresh=0.213
+cls_normalizer=1.0
+iou_normalizer=0.07
+iou_loss=ciou
+nms_kind=greedynms
+beta_nms=0.6
+max_delta=5