Recently, dense connections have attracted substantial attention in computer vision because they facilitate gradient flow and implicit deep supervision during training. Particularly, DenseNet, which connects each layer to every other layer in a feed-forward fashion, has shown impressive performances in natural image classification tasks. We propose HyperDenseNet, a 3D fully convolutional neural network that extends the definition of dense connectivity to multi-modal segmentation problems. Each imaging modality has a path, and dense connections occur not only between the pairs of layers within the same path, but also between those across different paths. This contrasts with the existing multi-modal CNN approaches, in which modeling several modalities relies entirely on a single joint layer (or level of abstraction) for fusion, typically either at the input or at the output of the network. Therefore, the proposed network has total freedom to learn more complex combinations between the modalities, within and in-between all the levels of abstraction, which increases significantly the learning representation. We report extensive evaluations over two different and highly competitive multi-modal brain tissue segmentation challenges, iSEG 2017 and MRBrainS 2013, with the former focusing on 6-month infant data and the latter on adult images. HyperDenseNet yielded significant improvements over many state-of-the-art segmentation networks, ranking at the top on both benchmarks. We further provide a comprehensive experimental analysis of features re-use, which confirms the importance of hyper-dense connections in multi-modal representation learning. Our code is publicly available at https://www.github.com/josedolz/HyperDenseNet.
翻译:最近,密集的连接在计算机视野中引起了大量关注,因为它们有利于梯度流和在培训过程中进行隐含的深度监督。 特别是,DenseNet(DenseNet)将每个层与其它层连接起来,在自然图像分类任务中表现出令人印象深刻的性能。 我们提议HerperDenseNet(3D全演化神经网络),将密集连接的定义扩大到多模式分割问题。每个成像模式都有一条路径,不仅在同一路径的层之间,而且在不同路径的层之间也存在密集的连接。这与现有的多模式CNN(M)方法形成对照,其中若干模式完全依赖单一的联合层(或抽象程度)来进行组合,通常是投入或网络输出。因此,拟设的3D全局性神经网络完全可以自由地了解这些模式之间更为复杂的组合,在各种抽象层次内和之间,这大大增加了学习代表性。我们报告对两种不同和高度竞争性的多模式脑组织分割的挑战进行了广泛的评价,即iSEG-2017和MRBRAINS(2013)的模型完全依赖于单一的联合层层(或抽象),在网络的单一联合层结构中,在6个月的Syal-ROD-deal-deal-deal-deal-deal-deal-deal-real-real-real-deal-deal-deal-deal-real-real-real-real-real-real-real-real-real-real-real-real-real-real-regy-real-real-real-regy-regy-real-real-de-real-al-de-de-de-de-de-de-de-de-de-de-de-de-real-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-de-