docs(yolo_v1): 新增文档

1. 训练日志
2. 架构解析
3. 工程介绍

Author: zjZSTU

docs/index.md

@@ -1 +1,19 @@
-# YOLO_v1
+# YOLO_v1
+
+实现`YOLO_v1`算法
+
+## 数据集
+
+使用`3`类定位数据集，参考[[数据集]Image Localization Dataset](https://blog.zhujian.life/posts/a2d65e1.html)
+
+```
+{'cucumber': 63, 'mushroom': 61, 'eggplant': 62}
+```
+
+## 实现流程
+
+1. 创建训练数据集
+2. 定义`YoLo_v1`模型
+3. 定义损失函数`Multi-part Loss`
+4. 训练模型
+5. 计算`mAP`

docs/log.md

@@ -0,0 +1,266 @@
+
+# 日志
+
+## 训练参数
+
+* `S=7, B=2, C=3`
+* 缩放至`(448, 448)`，进行数据标准化处理
+* 优化器：`SGD`，学习率`1e-3`，动量大小`0.9`
+* 衰减器：每隔`4`轮衰减`4%`，学习因子`0.96`
+
+## 训练日志
+
+```
+$ python train.py 
+Epoch 0/49
+----------
+train Loss: 4.3550
+save model
+
+Epoch 1/49
+----------
+train Loss: 3.4803
+save model
+
+Epoch 2/49
+----------
+train Loss: 3.3921
+save model
+
+Epoch 3/49
+----------
+train Loss: 3.0650
+save model
+
+Epoch 4/49
+----------
+train Loss: 2.9081
+save model
+
+Epoch 5/49
+----------
+train Loss: 2.5893
+save model
+
+Epoch 6/49
+----------
+train Loss: 2.5640
+save model
+
+Epoch 7/49
+----------
+train Loss: 2.4905
+save model
+
+Epoch 8/49
+----------
+train Loss: 2.1913
+save model
+
+Epoch 9/49
+----------
+train Loss: 2.1152
+save model
+
+Epoch 10/49
+----------
+train Loss: 1.9428
+save model
+
+Epoch 11/49
+----------
+train Loss: 1.7271
+save model
+
+Epoch 12/49
+----------
+train Loss: 1.4372
+save model
+
+Epoch 13/49
+----------
+train Loss: 1.8185
+
+Epoch 14/49
+----------
+train Loss: 1.5460
+
+Epoch 15/49
+----------
+train Loss: 1.1692
+save model
+
+Epoch 16/49
+----------
+train Loss: 1.0540
+save model
+
+Epoch 17/49
+----------
+train Loss: 0.9028
+save model
+
+Epoch 18/49
+----------
+train Loss: 0.7702
+save model
+
+Epoch 19/49
+----------
+train Loss: 0.7176
+save model
+
+Epoch 20/49
+----------
+train Loss: 0.7485
+
+Epoch 21/49
+----------
+train Loss: 0.6307
+save model
+
+Epoch 22/49
+----------
+train Loss: 0.5581
+save model
+
+Epoch 23/49
+----------
+train Loss: 0.5320
+save model
+
+Epoch 24/49
+----------
+train Loss: 0.5893
+
+Epoch 25/49
+----------
+train Loss: 0.5185
+save model
+
+Epoch 26/49
+----------
+train Loss: 0.6156
+
+Epoch 27/49
+----------
+train Loss: 0.5096
+save model
+
+Epoch 28/49
+----------
+train Loss: 0.5403
+
+Epoch 29/49
+----------
+train Loss: 0.4653
+save model
+
+Epoch 30/49
+----------
+train Loss: 0.3850
+save model
+
+Epoch 31/49
+----------
+train Loss: 0.3609
+save model
+
+Epoch 32/49
+----------
+train Loss: 0.4063
+
+Epoch 33/49
+----------
+train Loss: 0.3349
+save model
+
+Epoch 34/49
+----------
+train Loss: 0.2629
+save model
+
+Epoch 35/49
+----------
+train Loss: 0.3319
+
+Epoch 36/49
+----------
+train Loss: 0.2790
+
+Epoch 37/49
+----------
+train Loss: 0.2487
+save model
+
+Epoch 38/49
+----------
+train Loss: 0.2325
+save model
+
+Epoch 39/49
+----------
+train Loss: 0.2146
+save model
+
+Epoch 40/49
+----------
+train Loss: 0.2087
+save model
+
+Epoch 41/49
+----------
+train Loss: 0.1626
+save model
+
+Epoch 42/49
+----------
+train Loss: 0.1446
+save model
+
+Epoch 43/49
+----------
+train Loss: 0.1372
+save model
+
+Epoch 44/49
+----------
+train Loss: 0.1260
+save model
+
+Epoch 45/49
+----------
+train Loss: 0.1231
+save model
+
+Epoch 46/49
+----------
+train Loss: 0.1232
+
+Epoch 47/49
+----------
+train Loss: 0.1475
+
+Epoch 48/49
+----------
+train Loss: 0.1226
+save model
+
+Epoch 49/49
+----------
+train Loss: 0.1000
+save model
+
+Training complete in 6m 28s
+```
+
+## 检测结果
+
+```
+compute mAP
+{'cucumber': 63, 'mushroom': 61, 'eggplant': 62}
+99.43% = cucumber AP 
+98.39% = eggplant AP 
+99.22% = mushroom AP 
+mAP = 99.01%
+```

docs/架构解析.md

@@ -0,0 +1,37 @@
+
+# 架构解析
+
+使用`Python`实现，相关源码位于`py`目录下
+
+```
+├── batch_detect.py
+├── data
+├── detector.py
+├── lib
+    ├── data
+    │   ├── __init__.py
+    │   └── parse_location.py
+    ├── __init__.py
+    ├── models
+    │   ├── basic_conv2d.py
+    │   ├── __init__.py
+    │   ├── location_dataset.py
+    │   ├── multi_part_loss.py
+    │   ├── __pycache__
+    │   └── yolo_v1.py
+    ├── __pycache__
+    ├── train.py
+    └── utils
+        ├── draw.py
+        ├── file.py
+        ├── __init__.py
+        ├── __pycache__
+        ├── util.py
+        └── voc_map.py
+└── models
+```
+
+* 数据集操作：`lib/data/parse_location.py`
+* 模型定义：`lib/models`
+* 训练文件：`lib/train.py`
+* 批量测试：`batch_detect.py`

mkdocs.yml

@@ -29,4 +29,6 @@ extra_javascript:
     - 'https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.0/MathJax.js?config=TeX-MML-AM_CHTML'
 # 导航
 nav:
-    - Home: index.md
+    - Home: index.md
+    - 架构解析: '架构解析.md'
+    - log: log.md