How Does Topology Influence Gradient Propagation and Model Performance of Deep Networks With DenseNet-Type Skip Connections?

Kartikeya Bhardwaj, Guihong Li, Radu Marculescu; Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2021, pp. 13498-13507

Abstract


DenseNets introduce concatenation-type skip connections that achieve state-of-the-art accuracy in several computer vision tasks. In this paper, we reveal that the topology of the concatenation-type skip connections is closely related to the gradient propagation which, in turn, enables a predictable behavior of DNNs' test performance. To this end, we introduce a new metric called NN-Mass to quantify how effectively information flows through DNNs. Moreover, we empirically show that NN-Mass also works for other types of skip connections, e.g., for ResNets, Wide-ResNets (WRNs), and MobileNets, which contain addition-type skip connections (i.e., residuals or inverted residuals). As such, for both DenseNet-like CNNs and ResNets/WRNs/MobileNets, our theoretically grounded NN-Mass can identify models with similar accuracy, despite having significantly different size/compute requirements. Detailed experiments on both synthetic and real datasets (e.g., MNIST, CIFAR-10, CIFAR-100, ImageNet) provide extensive evidence for our insights. Finally, the closed-form equation of our NN-Mass enables us to design significantly compressed DenseNets (for CIFAR-10) and MobileNets (for ImageNet) directly at initialization without time-consuming training and/or searching.

Related Material


[pdf] [supp] [arXiv]
[bibtex]
@InProceedings{Bhardwaj_2021_CVPR, author = {Bhardwaj, Kartikeya and Li, Guihong and Marculescu, Radu}, title = {How Does Topology Influence Gradient Propagation and Model Performance of Deep Networks With DenseNet-Type Skip Connections?}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month = {June}, year = {2021}, pages = {13498-13507} }