全国人工智能大赛复赛方案 global feature部分,没进决赛懒得整理代码了(摊手)
-
Install dependencies:
- pytorch>=0.4
- torchvision
- ignite=0.1.2 (Note: V0.2.0 may result in an error)
- yacs
-
Bag of tricks
- Warm up learning rate
- Random erasing augmentation
- Label smoothing
- Last stride
- BNNeck
- Center loss
- Focal loss
- OSM Loss
- GCNet
-
初赛pretrained-weights 提取码: kmn2 copy to folder 'market1501'
-
训练方法
1.将数据集train 与 train_list.txt放置在dataset文件夹下,运行change_name.py,预处理数据集,将query 解压在 query文件夹下,gallery 解压在 bounding_box_test文件夹下。
2.运行train_resnet152.sh 训练模型 模型结果及模型权重保存在market1501下。