Keywords: zero-shot learning, normalization, continual learning, initialization
Abstract: Normalization techniques have proved to be a crucial ingredient of successful training in a traditional supervised learning regime. However, in the zero-shot learning (ZSL) world, these ideas have received only marginal attention. This work studies normalization in ZSL scenario from both theoretical and practical perspectives. First, we give a theoretical explanation to two popular tricks used in zero-shot learning: normalize+scale and attributes normalization and show that they help training by preserving variance during a forward pass. Next, we demonstrate that they are insufficient to normalize a deep ZSL model and propose Class Normalization (CN): a normalization scheme, which alleviates this issue both provably and in practice. Third, we show that ZSL models typically have more irregular loss surface compared to traditional classifiers and that the proposed method partially remedies this problem. Then, we test our approach on 4 standard ZSL datasets and outperform sophisticated modern SotA with a simple MLP optimized without any bells and whistles and having ~50 times faster training speed. Finally, we generalize ZSL to a broader problem — continual ZSL, and introduce some principled metrics and rigorous baselines for this new setup. The source code is available at https://github.com/universome/class-norm.
One-sentence Summary: We develop theoretical understanding of signal normalization inside zero-shot learning models, propose a novel normalization scheme and use it to achieve SotA ZSL performance with a simple MLP
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics
Data: [AwA](https://paperswithcode.com/dataset/awa-1), [AwA2](https://paperswithcode.com/dataset/awa2-1)