Multi-Scale Dense Networks for Resource Efficient Image Classification
Multi-Scale Dense Networks for Resource Efficient Image Classification
复制标题
DOI：
--
复制
发表时间：
2017-03
期刊：
影响因子：
--
通讯作者：
Gao Huang;Danlu Chen;Tianhong Li;Felix Wu;L. Maaten;Kilian Q. Weinberger
中科院分区：
文献类型：
--
作者：
Gao Huang;Danlu Chen;Tianhong Li;Felix Wu;L. Maaten;Kilian Q. Weinberger
在本文中，我们研究图像分类与计算资源的限制，在测试时间。两个这样的设置是：1.随时分类，其中网络对测试示例的预测逐步更新，便于在任何时间输出预测;以及2.预算批量分类，其中固定量的计算可用于对一组示例进行分类，这些示例可能在“更容易”和“更难”的输入之间不均匀地花费。与大多数先前的工作相比，例如流行的Viola和Jones算法，我们的方法基于卷积神经网络。我们用不同的资源需求训练多个分类器，我们在测试时自适应地应用这些分类器。为了最大限度地重用分类器之间的计算，我们将它们作为早期退出合并到单个深度卷积神经网络中，并将它们与密集连接进行互连。为了促进早期的高质量分类，我们使用了一个二维的多尺度网络架构，在整个网络中保持粗和细级别的功能。三个图像分类任务的实验表明，我们的框架大大提高了现有的国家的最先进的在这两种设置。
In this paper we investigate image classification with computational resource limits at test time. Two such settings are: 1. anytime classification, where the network's prediction for a test example is progressively updated, facilitating the output of a prediction at any time; and 2. budgeted batch classification, where a fixed amount of computation is available to classify a set of examples that can be spent unevenly across "easier" and "harder" inputs. In contrast to most prior work, such as the popular Viola and Jones algorithm, our approach is based on convolutional neural networks. We train multiple classifiers with varying resource demands, which we adaptively apply during test time. To maximally re-use computation between the classifiers, we incorporate them as early-exits into a single deep convolutional neural network and inter-connect them with dense connectivity. To facilitate high quality classification early on, we use a two-dimensional multi-scale network architecture that maintains coarse and fine level features all-throughout the network. Experiments on three image-classification tasks demonstrate that our framework substantially improves the existing state-of-the-art in both settings.