English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Continual learning involves training neural networks incrementally for newtasks while retaining the knowledge of previous tasks. However, efficientlyfine-tuning the model for sequential tasks with minimal computational resourcesremains a challenge. In this paper, we propose Task Incremental ContinualLearning (TI-CL) of audio classifiers with both parameter-efficient andcompute-efficient Audio Spectrogram Transformers (AST). To reduce the trainableparameters without performance degradation for TI-CL, we compare severalParameter Efficient Transfer (PET) methods and propose AST with ConvolutionalAdapters for TI-CL, which has less than 5% of trainable parameters of the fullyfine-tuned counterparts. To reduce the computational complexity, we introduce anovel Frequency-Time factorized Attention (FTA) method that replaces thetraditional self-attention in transformers for audio spectrograms. FTA achievescompetitive performance with only a factor of the computations required byGlobal Self-Attention (GSA). Finally, we formulate our method for TI-CL, calledAdapter Incremental Continual Learning (AI-CL), as a combination of the"parameter-efficient" Convolutional Adapter and the "compute-efficient" FTA.Experiments on ESC-50, SpeechCommandsV2 (SCv2), and Audio-Visual Event (AVE)benchmarks show that our proposed method prevents catastrophic forgetting inTI-CL while maintaining a lower computational budget.

Adapter Incremental Continual Learning of Efficient Audio Spectrogram Transformers