@inproceedings{Zhang:2024,
author = {Zhang, D and Williams, M and Toni, F},
publisher = {AAAI},
title = {Targeted activation penalties help CNNs ignore spurious signals},
url = {http://hdl.handle.net/10044/1/108884},
year = {2024}
}

Download

TY  - CPAPER
AB  - Neural networks (NNs) can learn to rely on spurious signals in the training data, leading to poor generalisation. Recent methods tackle this problem by training NNs with additional ground-truth annotations of such signals. These methods may, however, let spurious signals re-emerge in deep convolutional NNs (CNNs). We propose Targeted Activation Penalty (TAP), a new method tackling the same problem by penalising activations to control the re-emergence of spurious signals in deep CNNs, while also lowering training times and memory usage. In addition, ground-truth annotations can be expensive to obtain. We show that TAP still works well with annotations generated by pre-trained models as effective substitutes of ground-truth annotations. We demonstrate the power of TAP against two state-of-the-art baselines on the MNIST benchmark and on two clinical image datasets, using four different CNN architectures.
AU  - Zhang,D
AU  - Williams,M
AU  - Toni,F
PB  - AAAI
PY  - 2024///
SN  - 2159-5399
TI  - Targeted activation penalties help CNNs ignore spurious signals
UR  - http://hdl.handle.net/10044/1/108884
ER  - 

Download