Research And Implementation On Open Set Image Classification Algorithm Based On Domain Adaptation

With the widespread use of deep learning in image,text,and speech domains,high-quality datasets have become a very scarce and precious resource.However,collecting enough labeled data for each deep learning application scenario is a time-consuming and costly process.Domain adaptive techniques in transfer learning can migrate or reuse knowledge from a relevant source domain with existing high-quality datasets and free from the dependence on labeled or supervised information of data in the target domain scenario.Most existing domain adaptation frameworks were based on the strict assumption that different domains share the same label space,which is too idealized for real-world applications.In this paper,we examine a realistic case of open set domain adaptation with partially shared classes in two domains and consider the negative transfer brought by the unknown class from the target domain.Therefore,how to distinguish the unknown class from the known ones plays a key role.To achieve this goal,we propose a new approach called Open Distribution Control Adaptation Network(OpenDCAN).Inspired by the idea of Out-of-Distribution Detection,our method tries to explicitly model the distribution of the known source samples to differentiate it from unknown class in probabilistic distribution,which is different from existing methods.We use OpenBP,an open set domain adaptation framework,to provide a rough decision boundary between known and unknown classes.Then we further boost the performance by modifying the recently proposed Adaptive Feature Norm algorithm for the open set setting.If the probabilistic distribution of source samples is modeled,a likelihood estimate can easily detect unknown class without requiring any unknown samples in the source domain to get the concept of unknown.Experiments on representative benchmark datasets demonstrate the effectiveness of our approach.Our method can identify unknown classes more simply and effectively.In terms of research landing,we designed and implemented an open-set image recognition system based on open-set domain adaptation algorithm with online transfer-learning and convenient data annotation function.