Mixed Link Networks
This work addresses the need for more efficient neural network architectures in computer vision, though it is incremental as it builds on existing ResNet and DenseNet frameworks.
The paper tackles the problem of combining the advantages of ResNet and DenseNet architectures by proposing Mixed Link Networks (MixNet), which integrate inner and outer link modules, achieving superior parameter efficiency on datasets like CIFAR-10/100, SVHN, and ImageNet.
Basing on the analysis by revealing the equivalence of modern networks, we find that both ResNet and DenseNet are essentially derived from the same "dense topology", yet they only differ in the form of connection -- addition (dubbed "inner link") vs. concatenation (dubbed "outer link"). However, both two forms of connections have the superiority and insufficiency. To combine their advantages and avoid certain limitations on representation learning, we present a highly efficient and modularized Mixed Link Network (MixNet) which is equipped with flexible inner link and outer link modules. Consequently, ResNet, DenseNet and Dual Path Network (DPN) can be regarded as a special case of MixNet, respectively. Furthermore, we demonstrate that MixNets can achieve superior efficiency in parameter over the state-of-the-art architectures on many competitive datasets like CIFAR-10/100, SVHN and ImageNet.