Incremental Object Learning From Contiguous Views

• Published in: 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) pp. 8769 - 8778
• Main Authors: Stojanov, Stefan, Mishra, Samarth, Thai, Ngoc Anh, Dhanda, Nikhil, Humayun, Ahmad, Yu, Chen, Smith, Linda B., Rehg, James M.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01-06-2019
• Subjects: Categorization; Datasets and Evaluation; Deep Learning; Image and Video Synthesis; Recognition: Detection; Retrieval
• ISSN: 2575-7075
• DOI: 10.1109/CVPR.2019.00898

In this work, we present CRIB (Continual Recognition Inspired by Babies), a synthetic incremental object learning environment that can produce data that models visual imagery produced by object exploration in early infancy. CRIB is coupled with a new 3D object dataset, Toys-200, that contains 200 unique toy-like object instances, and is also compatible with existing 3D datasets. Through extensive empirical evaluation of state-of-the-art incremental learning algorithms, we find the novel empirical result that repetition can significantly ameliorate the effects of catastrophic forgetting. Furthermore, we find that in certain cases repetition allows for performance approaching that of batch learning algorithms. Finally, we propose an unsupervised incremental learning task with intriguing baseline results.