Can selfless learning improve accuracy of a single classification task?

Roy, S. and Sau, B.B (2021) Can selfless learning improve accuracy of a single classification task? In: 2021 IEEE Winter Conference on Applications of Computer Vision (WACV), 5 January 2021 through 9 January 2021, Virtual, Online.

Full text not available from this repository.


The human brain has billions of neurons. However, we perform tasks using only a few concurrently active neurons. Moreover, an activated neuron inhibits the activity of its neighbors. Selfless Learning exploits these neurobiological principles to solve the problem of catastrophic forgetting in continual learning. In this paper, we ask a basic question: can the selfless learning idea be used to improve the accuracy of deep convolutional networks on a single classification task? To achieve this goal, we introduce two regularizers and formulate a curriculum learning-esque strategy to effectively enforce these regularizers on a network. This has resulted in significant gains over vanilla cross-entropy training. Moreover, we have shown that our method can be used in conjunction with other popular learning paradigms like curriculum learning.

[error in script]
IITH Creators:
IITH CreatorsORCiD
-NA-, -NA--NA-
Item Type: Conference or Workshop Item (Paper)
Uncontrolled Keywords: Active neurons; Catastrophic forgetting; Classification tasks, Continual learning, Convolutional networks, Cross entropy, Human brain, Learning paradigms, Regularizer
Subjects: Computer science
Divisions: Department of Computer Science & Engineering
Depositing User: Mrs Haseena VKKM
Date Deposited: 23 Nov 2021 06:09
Last Modified: 23 Nov 2021 06:09
Publisher URL:
Related URLs:

    Actions (login required)

    View Item View Item
    Statistics for RAIITH ePrint 8991 Statistics for this ePrint Item