Variational Multi-Prototype Encoder for Object Recognition Using Multiple Prototype Images

• Published in: IEEE access Vol. 10; pp. 19586 - 19598
• Main Authors: Kang, Jun Seok, Ahn, Sang Chul
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 2022; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Coders; Deep learning; embedding space; Feature extraction; image classification; Labeling; Learning; Neural networks; Object recognition; Performance degradation; Perturbation methods; prototype learning; Prototypes; Prototyping; Task analysis; Training; variational encoder
• ISSN: 2169-3536; 2169-3536
• DOI: 10.1109/ACCESS.2022.3151856

In the recent research of Variational Prototyping-Encoder (VPE), the problem of classifying 2D flat objects of the unseen class has been addressed. VPE solves this problem by pre-learning the image translation task from real-world object images to their corresponding prototype images as a meta-task. VPE uses a single prototype for each object class. However, in general, a single prototype is not sufficient to represent a generic object class because the appearance can change significantly according to viewpoints and other factors. In this case, using VPE and a single prototype for each class in training can result in overfitting or performance degradation. One solution may be the use of multiple prototypes. However, this also requires costly sub-labeling for dividing the input class into smaller classes and assigning a prototype to each. Therefore, we propose a new learning method, the variational multi-prototype encoder (VaMPE), which can overcome the limitations of VPE and use multiple prototypes for each object class. The proposed method does not require additional sub-labeling other than simply adding multiple prototypes to each class. Through various experiments, we demonstrate that the proposed method outperforms VPE.