[论文解读] Learning a Hierarchical Compositional Shape Vocabulary for Multi-class Object Representation

Sanja Fidler, Marko Boben+1 more|arXiv (Cornell University)|Aug 23, 2014

Advanced Image and Video Retrieval Techniques参考文献 10被引用 20

一句话总结

该论文提出了一种无监督的自底向上框架，用于从定向轮廓片段中学习分层组合形状词汇表，通过递归组合形成越来越复杂的、特定类别的形状构型。该方法实现了最先进水平的检测性能，词汇表大小和推理复杂度呈对数增长，实现了可扩展的多类别目标识别，具备快速推理和短时训练的特点。

ABSTRACT

Hierarchies allow feature sharing between objects at multiple levels of representation, can code exponential variability in a very compact way and enable fast inference. This makes them potentially suitable for learning and recognizing a higher number of object classes. However, the success of the hierarchical approaches so far has been hindered by the use of hand-crafted features or predetermined grouping rules. This paper presents a novel framework for learning a hierarchical compositional shape vocabulary for representing multiple object classes. The approach takes simple contour fragments and learns their frequent spatial configurations. These are recursively combined into increasingly more complex and class-specific shape compositions, each exerting a high degree of shape variability. At the top-level of the vocabulary, the compositions are sufficiently large and complex to represent the whole shapes of the objects. We learn the vocabulary layer after layer, by gradually increasing the size of the window of analysis and reducing the spatial resolution at which the shape configurations are learned. The lower layers are learned jointly on images of all classes, whereas the higher layers of the vocabulary are learned incrementally, by presenting the algorithm with one object class after another. The experimental results show that the learned multi-class object representation scales favorably with the number of object classes and achieves a state-of-the-art detection performance at both, faster inference as well as shorter training times.

研究动机与目标

• 开发一种可扩展的多类别目标表示方法，无需人工标注即可捕捉复杂形状结构。
• 通过引入分层、组合式形状建模，解决平面词袋模型的局限性。
• 在多个抽象层次上实现跨类别特征共享，以提升泛化能力和效率。
• 以自底向上、统计化的方式学习形状词汇表，最大限度减少人工干预，避免手工设计的特征或固定分组规则。

提出的方法

• 该方法将简单的定向轮廓片段作为基础层级，并识别其频繁的空间配置。
• 通过将低层级部件基于空间关系（建模为高斯分布）进行组合，递归构建出越来越复杂的分层结构。
• 低层在所有类别上联合训练，以捕捉通用形状结构；高层则按类别逐步学习。
• 分析窗口大小随层级递增，空间分辨率递减，实现多尺度形状建模。
• 每个构型均为生成式概率模型，可捕捉前一层部件的分布，支持形变建模。
• 该框架采用分层、自底向上的学习过程，可高效扩展至更多类别。

实验结果

研究问题

• RQ1 能否从简单轮廓片段出发，无监督地学习分层、组合式形状词汇表，以表示多个类别？
• RQ2 与平面表示相比，分层组合如何提升多类别目标检测中的泛化能力和推理效率？
• RQ3 跨类别的共享特征在多大程度上可减少词汇表大小和训练时间，同时保持高检测准确率？
• RQ4 该方法在类别数量增加时是否能有效扩展，同时保持快速推理和紧凑表示？

主要发现

• 该方法在多个类别上实现了最先进水平的检测性能，包括瓶子、长颈鹿、水杯和汽车变体等。
• 推理时间随类别数量呈对数增长，显著优于平面方法。
• 词汇表大小在低层呈对数增长，即使类别数增加也能实现可扩展表示。
• 模型检测准确率高：对于汽车（前视图）在0.4 FPPI下达到97.5%的检测率，对牛的检测率为96.9%。
• 该框架展现出强大的泛化能力，在人脸检测中EER处达到93.0%的召回率，在行人检测中达到85.0%。
• 该方法实现快速训练与推理，无需人工部件标注或预设分组规则。