The self-ensembling methods have achieved amazing performance for semi-supervised representation learning and domain adaptation. However, the disadvantage of these methods is that the teacher network is tightly coupled with the student network, which limits the descriptive ability of the self-ensembling model. To overcome the coupling effect between the teacher network and the student network, we propose a novel Triplet Ensemble Student-Teacher (TEST) model for unsupervised person re-identification, which consists of one teacher network TT and two student networks S1S1 and S2S2 . Similar to the traditional self-ensembling model, the student network S1S1 is applied to update the teacher network TT . Furthermore, a closed-loop learning mechanism is built in the TEST model by imposing an ensemble consistent constraint between TT and S2S2 , and performing a heterogeneous co-teaching procedure between S1S1 and S2S2 . With the closed-loop learning mechanism, the TEST model can loosen the constraint between the teacher TT and the student S1S1 , and enhance the descriptive ability of S1S1 . Besides, the knowledge exchange between S1S1 and S2S2 can ensure that the two student networks can elegantly deal with the noisy labels and avoid coupling. By training the TEST model with the clustering-generated pseudo labels, we can achieve effective and robust representation learning for unsupervised person re-identification. The evaluations on three widely-used benchmarks show that our approach can achieve significant performance compared with state-of-the-art methods.

中文翻译：

---

测试：用于无监督人员重新识别的三元组学生-教师模型


自集成方法在半监督表示学习和领域适应方面取得了惊人的性能。然而这些方法的缺点是教师网络与学生网络紧密耦合，限制了自集成模型的描述能力。为了克服教师网络和学生网络之间的耦合效应，我们提出了一种用于无监督人员重新识别的三元组集成学生-教师（TEST）模型，该模型由一个教师网络 TT 和两个学生网络 S1S1 和 S2S2 组成。与传统的自集成模型类似，学生网络 S1S1 用于更新教师网络 TT 。此外，通过在TT和S2S2之间施加集成一致性约束，并在S1S1和S2S2之间执行异构协同教学过程，在TEST模型中建立了闭环学习机制。通过闭环学习机制，TEST模型可以放松教师TT和学生S1S1之间的约束，增强S1S1的描述能力。此外，S1S1和S2S2之间的知识交换可以确保两个学生网络能够优雅地处理噪声标签并避免耦合。通过使用聚类生成的伪标签训练 TEST 模型，我们可以实现有效且鲁棒的表示学习，以进行无监督的人员重新识别。对三个广泛使用的基准的评估表明，与最先进的方法相比，我们的方法可以实现显着的性能。