A compositional parts based model for object recognition

基于组合部件的物体识别模型

• 批准号: 1725305
• 金额: --
• 依托单位: University of Birmingham
• 依托单位国家: 英国
• 项目类别: Studentship
• 财政年份: 2015
• 资助国家: 英国
• 起止时间: 2015 至 无数据
• 项目状态: 已结题
• 来源: https://gtr.ukri.org/projects?ref=studentship-1725305
• 关键词: compositional; parts; based; model; object

Deep Learning architectures based on Convolutional Neural Networks (CNNs) are the state of the art technique for object recognition and segmentation. However, CNNs require millions of labelled images to reach this performance. They also suffer from a lack of interpretability making it difficult to understand why some images are misclassified. Recent work on adversarial attacks on these types of systems has highlighted this problem. Alternatives to CNNs such as parts based compositional models rivalled or surpassed CNNs performance in the past before the advances in network architecture and large-scale GPU training allowed CNNs to reach their current performance. Parts based compositional models can overcome some of the weaknesses of CNNs, but themselves suffer from poor scalability on large-scale datasets. A compositional model built using neural network components, but forming an interpretable compositional graph would in principle combine the advantages of both approaches in a scaleable, interpretable model.

基于卷积神经网络 (CNN) 的深度学习架构是最先进的对象识别和分割技术。然而，CNN 需要数百万张标记图像才能达到这种性能。它们还缺乏可解释性，因此很难理解为什么某些图像被错误分类。最近针对此类系统的对抗性攻击的研究凸显了这个问题。在网络架构和大规模 GPU 训练的进步使 CNN 达到当前性能之前，CNN 的替代方案（例如基于部件的组合模型）在过去可以与 CNN 的性能相媲美或超越。基于部件的组合模型可以克服 CNN 的一些弱点，但其本身在大规模数据集上的可扩展性较差。使用神经网络组件构建的组合模型，但形成可解释的组合图，原则上将在可扩展、可解释的模型中结合这两种方法的优点。