Details

Autor(en) / Beteiligte

• Liu, Mengchen
• Shi, Jiaxin
• Li, Zhen
• Li, Chongxuan
• Zhu, Jun
• Liu, Shixia

Titel

Towards Better Analysis of Deep Convolutional Neural Networks

Ist Teil von

• IEEE transactions on visualization and computer graphics, 2017-01, Vol.23 (1), p.91-100

Ort / Verlag

United States: IEEE

Erscheinungsjahr

2017

Quelle

IEEE Xplore

Beschreibungen/Notizen

• Deep convolutional neural networks (CNNs) have achieved breakthrough performance in many pattern recognition tasks such as image classification. However, the development of high-quality deep models typically relies on a substantial amount of trial-and-error, as there is still no clear understanding of when and why a deep model works. In this paper, we present a visual analytics approach for better understanding, diagnosing, and refining deep CNNs. We formulate a deep CNN as a directed acyclic graph. Based on this formulation, a hybrid visualization is developed to disclose the multiple facets of each neuron and the interactions between them. In particular, we introduce a hierarchical rectangle packing algorithm and a matrix reordering algorithm to show the derived features of a neuron cluster. We also propose a biclustering-based edge bundling method to reduce visual clutter caused by a large number of connections between neurons. We evaluated our method on a set of CNNs and the results are generally favorable.