arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1602.02660cs.LGcs.CVcs.NE
收藏

利用卷积神经网络中的循环对称性

Exploiting Cyclic Symmetry in Convolutional Neural Networks

Google DeepMind(谷歌DeepMind)

Sander Dieleman, Jeffrey De Fauw, Koray Kavukcuoglu

摘要:

许多类别的图像表现出旋转对称性。有时使用数据增强来训练卷积神经网络以利用这一点,但它们仍然需要从数据中学习旋转等方差特性。将这些属性编码到网络架构中,就像我们已经习惯于使用卷积层进行平移等方差一样,可以通过减轻模型的学习来更有效地使用参数预算。我们引入了四种操作,它们可以作为层插入到神经网络模型中,并且可以组合起来使这些模型与旋转部分等变。它们还支持在不同方向上共享参数。我们评估了这些架构修改对三个数据集的影响,这些数据集表现出旋转对称性,并在较小的模型中表现出更好的性能。

英文摘要:

Many classes of images exhibit rotational symmetry. Convolutional neural networks are sometimes trained using data augmentation to exploit this, but they are still required to learn the rotation equivariance properties from the data. Encoding these properties into the network architecture, as we are already used to doing for translation equivariance by using convolutional layers, could result in a more efficient use of the parameter budget by relieving the model from learning them. We introduce four operations which can be inserted into neural network models as layers, and which can be combined to make these models partially equivariant to rotations. They also enable parameter sharing across different orientations. We evaluate the effect of these architectural modifications on three datasets which exhibit rotational symmetry and demonstrate improved performance with smaller models.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑