OverFeat: Integrated Recognition, Localization and Detection using Convolutional Networks
Pierre Sermanet David Eigen Xiang Zhang Michael Mathieu Rob Fergus Yann LeCun Courant Institute of Mathematical Sciences, New York University 719 Broadway, 12th Floor, New York, NY 10003
Abstract
We present an integrated framework for using Convolutional Networks for classification, localization and detection. We show how a multiscale and sliding window approach can be efficiently implemented within a ConvNet. We also introduce a novel deep learning approach to localization by learning to predict object boundaries. Bounding boxes are then accumulated rather than suppressed in order to increase detection confidence. We show that different tasks can be learned simultaneously using a single shared network. This integrated framework is the winner of the localization task of the ImageNet Large Scale Visual Recognition Challenge 2013 (ILSVRC2013) and obtained very competitive results for the detection and classifications tasks. In post-competition work, we establish a new state of the art for the detection task. Finally, we release a feature extractor from our best model called OverFeat.
中文速览
卷积神经网络在图像分类任务上已取得亮眼成绩，但如何让同一套网络同时完成"图像里有什么"（分类）、"它在哪里"（定位）和"找出所有目标"（检测）三件事，仍缺乏一个统一的解决方案。论文提出了一个名为 OverFeat 的集成框架，核心思路是：用多尺度滑动窗口在整张图上密集运行卷积网络，同时让网络不仅预测类别概率，还直接回归目标边界框；随后将各位置、各尺度预测到的边界框进行"累积叠加"而非传统的非极大值抑制，以此提升检测置信度，整个流程共用一套网络权重、端到端训练。该框架在 ILSVRC 2013 竞赛中赢得定位任务冠军，并在赛后实验中刷新了检测任务的最优成绩，同时作者还开源了特征提取器 OverFeat，为后续视觉研究提供了强力的预训练基础。这项工作的重要性在于，它首次清晰地展示了单个卷积网络可以统一处理分类、定位与检测，为后来目标检测方法的演进奠定了基础。
原文 arXiv:1312.6229;中英对照 + 大白话阅读 https://aha.fim.ai/paper/1312.6229v4