RSRuibin Song

Ruibin Song

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Ruibin Song has a total of 12 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Transferring Face Recognition Techniques to Entomology: An ArcFace and ResNet Approach for Improving Dragonfly Classification

    by Zhong Li, Shaoyan Pu, Jingsheng Lu, Ruibin Song, Haomiao Zhang, Xuemei Lu, Yanan Wang

    Dragonfly classification is crucial for biodiversity conservation. Traditional taxonomic approaches require extensive training and experience, limiting their efficiency. Computer vision offers promising solutions for dragonfly taxonomy. In this study, we adapt the face recognition algorithms for the classification of dragonfly species, achieving efficient recognition of categories with extremely small differences between classes. Meanwhile, this method can also reclassify data that were incorrectly labeled. The model is mainly built based on the classic face recognition algorithm (ResNet50+ArcFace), and ResNet50 is used as the comparison algorithm for model performance. Three datasets with different inter-class data distributions were constructed based on two dragonfly image data sources: Data1, Data2 and Data3. Ultimately, our model achieved Top1 accuracy rates of 94.3%, 85.7%, and 90.2% on the three datasets, surpassing ResNet50 by 0.6, 1.5, and 1.6 percentage points, respectively. Under the confidence thresholds of 0.7, 0.8, 0.9, and 0.95, the Top1 accuracy rates on the three datasets were 96.0%, 97.4%, 98.7%, and 99.2%, respectively. In conclusion, our research provides a novel approach for species classification. Furthermore, it can calculate the similarity between classes while predicting categories, thereby offering the potential to provide technical support for biological research on the similarity between species.


  • Epiophlebia laidlawi daliensis, a new subspecies from Hengduan Mountains in Yunnan (Insecta, Odonata, Epiophlebiidae)

    by Guo-Hui Yang, Yun-Yang Wei, Zi-Feng Li, Rui-Bin Song, Yu-Chuan Wen, Xiao-Long Li, Dong-Dong Zi, Xue-Mei Lyu, Hao-Miao Zhang

    A new subspecies Epiophlebia laidlawi daliensis Zhang, Yang, Wei & Lyu, subsp. nov. is described from adult and larval specimens collected at Mt Cangshan in the Hengduan Mountains, Dali Bai Autonomous Prefecture, Yunnan, China. The holotype male was collected by Hao-miao Zhang and Rui-bin Song on 22 May 2022. Both sexes and larva of the new subspecies are compared with those of Epiophlebia l. laidlawi Tillyard, 1921 and E. superstes (Selys, 1889). A total of 15 sequences of ITS1 and 17 sequences of ITS2 were analysed and used for the phylogenetic comparisons. According to this analysis, E. l. daliensis subsp. nov. and E. l. laidlawi share a very close phylogenetic relationship, while E. l. daliensis, subsp. nov. and E. superstes are genetically more distant from one another than are E. l. laidlawi and E. superstes. Both neighbor-joining (NJ) and maximum likelihood (ML) methods produced consistent phylogenetic trees, with all E. l. daliensis specimens clustering with E. l. laidlawi. The existence of the named ‘species’ E. sinensis Li & Nel, 2012 (which in molecular analyses is retrieved within E. superstes) is considered doubtful, hence it is not considered here. Its status will be evaluated in another publication.


  • Open Access

    Transferring Face Recognition Techniques to Entomology: An ArcFace and ResNet Approach for Improving Dragonfly Classification

    by Zhong Li, Shaoyan Pu, Jingsheng Lu, Ruibin Song, Haomiao Zhang, Xuemei Lu, Yanan Wang

    Dragonfly classification are crucial for biodiversity conservation. Traditional taxonomic approaches require extensive training and experience, limiting their efficiency. Computer vision offers promising solutions for dragonfly taxonomy.In this study, we adapt the face recognition algorithms for the classification of dragonfly species, achieving efficient recognition of categories with extremely small differences between classes. Meanwhile, this method can also reclassify data that was incorrectly labeled. The model is mainly built based on the classic face recognition algorithm (ResNet50+ArcFace), and ResNet50 is used as the comparison algorithm for model performance. Three datasets with different inter-class data distributions were constructed based on two dragonfly image data sources.: Data1, Data2 and Data3. Ultimately, our model achieved Top1 accuracy rates of 94.3%, 85.7%, and 90.2% on the three datasets, surpassing ResNet50 by 0.6, 1.5, and 1.6 percentage points, respectively. Under the confidence thresholds of 0.7, 0.8, 0.9, and 0.95, the Top1 accuracy rates on the three datasets were 96.0%, 97.4%, 98.7%, and 99.2%, respectively. In conclusion, our research provides a novel approach for species classification. Furthermore, it can calculate the similarity between classes while predicting categories, thereby offering the potential to provide technical support for biological research on the similarity between species.