Name
Abstract | ||
|---|---|---|
Neural architectures are the foundation for improving performance of deep neural networks (DNNs). This paper presents deep compositional grammatical architectures which harness the best of two worlds: grammar models and DNNs. The proposed architectures integrate compositionality and reconfigurability of the former and the capability of learning rich features of the latter in a principled way. We utilize AND-OR Grammar (AOG) [53, 71, 70] as network generator in this paper and call the resulting networks AOGNets. An AOGNet consists of a number of stages each of which is composed of a number of AOG building blocks. An AOG building block splits its input feature map into N groups along feature channels and then treat it as a sentence of N words. It then jointly realizes a phrase structure grammar and a dependency grammar in bottom-up parsing the "sentence" for better feature exploration and reuse. It provides a unified framework for the best practices developed in state-of-the-art DNNs. In experiments, AOGNet is tested in the ImageNet-1K classification benchmark and the MS-COCO object detection and segmentation benchmark. In ImageNet-1K, AOGNet obtains better performance than ResNet [21] and most of its variants, ResNeXt [63] and its attention based variants such as SENet [24], DenseNe[26] and DualPathNet [6]. AOGNet also obtains the best model interpretability score using network dissection [3]. AOGNet further shows better potential in adversarial defense. In MS-COCO, AOGNet obtains better performance than the ResNet and ResNeXt backbones in Mask R-CNN [20]. |
Year | DOI | Venue |
|---|---|---|
2019 | 10.1109/CVPR.2019.00638 | 2019 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2019) |
Field | DocType | ISSN |
Principle of compositionality,Interpretability,Computer science,Phrase structure grammar,Grammar,Dependency grammar,Artificial intelligence,Parsing,Deep learning,Sentence,Machine learning | Conference | 1063-6919 |
Citations | PageRank | References |
2 | 0.36 | 0 |
Authors | ||
3 | ||