BUAA-iCC at ImageCLEF 2015 Scalable Concept Image Annotation Challenge
Yunhong Wang, Jiaxin Chen, Ningning Liu, Li Zhang · CLEF (Working Notes) · 2015
In this working note, we mainly focus on the image anno- tation subtask of ImageCLEF 2015 challenge that BUAA-iCC research group participated. For this task, we flrstly explore textual similarity information between each test sample and predeflned concept. Subse- quently, two difierent kinds of semantic information are extracted from visual images: visual tags using generic object recognition classiflers and visual tags relevant to human being related concepts. For the former information, the visual tags are predicted by using deep convolutional neural network (CNN) and a set of support vector machines trained on ImageNet, and flnally transferred to textual information. For the latter visual information, human related concepts are extracted via face and facial attribute detection, and flnally transferred to similarity informa- tion by using manually designed mapping rules, in order to enhance the performance of annotating human related concepts. Meanwhile, a late fusion strategy is developed to incorporate aforementioned various kinds of similarity information. Results validate that the combination of the textual and visual similarity information and the adopted late fusion strategy could yield signiflcantly better performance.