We propose a neural network architecture that learns body part appearances
for soccer player re-identification. Our model consists of a two-stream network
(one stream for appearance map extraction and the other for body part map
extraction) and a bilinear-pooling layer that generates and spatially pools the
body part map. Each local feature of the body part map is obtained by a
bilinear mapping of the corresponding local appearance and body part
descriptors. Our novel representation yields a robust image-matching feature
map, which results from combining the local similarities of the relevant body
parts with the weighted appearance similarity. Our model does not require any
part annotation on the SoccerNet-V3 re-identification dataset to train the
network. Instead, we use a sub-network of an existing pose estimation network
(OpenPose) to initialize the part substream and then train the entire network
to minimize the triplet loss. The appearance stream is pre-trained on the
ImageNet dataset, and the part stream is trained from scratch for the
SoccerNet-V3 dataset. We demonstrate the validity of our model by showing that
it outperforms state-of-the-art models such as OsNet and InceptionNet.

我们提出了一种神经网络架构，用于学习足球运动员再识别的身体部位外观。我们的模型包括一个双流网络（一个用于外观地图提取，另一个用于身体部位地图提取）和一个双线性池化层，用于生成和空间汇聚身体部位地图。我们的模型通过将相应的局部外观和身体部位描述符进行双线性映射来获得身体部位地图的每个局部特征。我们的新颖表示产生了一个强大的图像匹配特征图，它由相关身体部位的局部相似性与加权外观相似性的组合结果得到。我们的模型不需要在 SoccerNet-V3 再识别数据库上对身体部位进行任何注释，而是使用现有姿态估计网络（OpenPose）的子网络来初始化部位子流，然后训练整个网络以最小化三元组损失。外观流使用 ImageNet 数据集进行预训练，而部位流从头开始训练 SoccerNet-V3 数据集。通过展示其优于 OsNet 和 InceptionNet 等最先进模型的性能，我们证明了我们模型的有效性。