As long as certain key features of an object are present in the test data, CNN's classify the test data as the object, disregarding features' relative spatial orientation to each other. This causes false positives. The lack of rotational invariance in CNN's would cause the network to ...
[10] E. Xi, S. Bing, and Y. Jin, ``Capsule network performance on complex data,'' Dec. 2017, arXiv:1712.03480v1. [Online]. Available:https://arxiv.org/abs/1712.03480v1 [11] D. Wang and Q. Liu, ``An optimization view on dynamic routing between capsules,'' presented at the 6t...
Capsule Network Performance on Complex Data - Xi, E., Bing, S. and Jin, Y. (2017) Accurate reconstruction of image stimuli from human fMRI based on the decoding model with capsule network architecture - Qiao, K., Zhang, C., Wang, L., Yan, B., Chen, J., Zeng, L. and Tong, L...
我们对于上述两个capsule模型设置了一组对比实验,是参数规模完全相同的普通卷积网络,stride设置也相同,在训练速度上,普通卷积网络快很多,但是在最终结果上,模型收敛后,只达到了65%的识别准确率,在这个角度上,capsule结构还是优于普通cnn的。 另有文章《Capsule Network Performance on Complex Data》对capsule在cifar10上...
However, CapsNet gets a poor performance on more complex datasets like CIFAR-10. To address this problem, we focus on the improvement of the original CapsNet from both the network structure and the dynamic routing mechanism. A new CapsNet architecture aiming at complex data called Capsule Network...
Xi, E., Bing, S. & Jin, Y. Capsule network performance on complex data.arXiv:1712.03480(arXiv preprint) (2017). Wang, D. & Liu, Q. An optimization view on dynamic routing between capsules (2018). Lenssen, J. E., Fey, M. & Libuschewski, P. Group equivariant capsule networks....
Jin Capsule Network Performance on Complex Data arXiv1712.03480v1 [stat.ML] (2017), pp. 1-7 View in ScopusGoogle Scholar Xia et al., 2018 C. Xia, C. Zhang, X. Yan, Y. Chang, P.S. Yu Zero-shot User Intent Detection via Capsule Neural Networks metharXiv 1809.00385v1 [cs.CL] (...
@GGWP同意这位同学说的，感觉最直观的解释就是参数太多了。CapsuleNet的理论很优雅，但是实现并不是很好...
cnn可以通过Max-pooling+BP来消除背景因素的影响，只留下物体的基本形状特征。我觉得如果要在这方面提升...
This work explores the performance of different capsule network architectures against simpler convolutional neural network architectures, in reproducing the classifications into the classes of unresolved, FRI, and FRII morphologies. We utilize images from a LOFAR survey which is the deepest, wide-area ...