Bilinear Convolutional Neural Networks for Fine-Grained Visual Recognition

• Published in: IEEE transactions on pattern analysis and machine intelligence Vol. 40; no. 6; pp. 1309 - 1322
• Main Authors: Tsung-Yu Lin, RoyChowdhury, Aruni, Maji, Subhransu
• Format: Journal Article
• Language: English
• Published: United States IEEE 01.06.2018; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Artificial neural networks; bilinear models; Birds; Computer architecture; Convolutional codes; convolutional networks; Feature extraction; Fine-grained recognition; Image classification; Image recognition; Neural networks; Object recognition; second order pooling; Source code; Texture recognition; texture representations; Visualization
• ISSN: 0162-8828; 1939-3539; 2160-9292; 1939-3539
• DOI: 10.1109/TPAMI.2017.2723400

We present a simple and effective architecture for fine-grained recognition called Bilinear Convolutional Neural Networks (B-CNNs). These networks represent an image as a pooled outer product of features derived from two CNNs and capture localized feature interactions in a translationally invariant manner. B-CNNs are related to orderless texture representations built on deep features but can be trained in an end-to-end manner. Our most accurate model obtains 84.1, 79.4, 84.5 and 91.3 percent per-image accuracy on the Caltech-UCSD birds [1], NABirds [2], FGVC aircraft [3], and Stanford cars [4] dataset respectively and runs at 30 frames-persecond on a NVIDIA Titan X GPU. We then present a systematic analysis of these networks and show that (1) the bilinear features are highly redundant and can be reduced by an order of magnitude in size without significant loss in accuracy, (2) are also effective for other image classification tasks such as texture and scene recognition, and (3) can be trained from scratch on the ImageNet dataset offering consistent improvements over the baseline architecture. Finally, we present visualizations of these models on various datasets using top activations of neural units and gradient-based inversion techniques. The source code for the complete system is available at http://vis-www.cs.umass.edu/bcnn.