Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Nested Hierarchical Transformer: Towards Accurate, Data-Efficient and Interpretable Visual Understanding

About

Hierarchical structures are popular in recent vision transformers, however, they require sophisticated designs and massive datasets to work well. In this paper, we explore the idea of nesting basic local transformers on non-overlapping image blocks and aggregating them in a hierarchical way. We find that the block aggregation function plays a critical role in enabling cross-block non-local information communication. This observation leads us to design a simplified architecture that requires minor code changes upon the original vision transformer. The benefits of the proposed judiciously-selected design are threefold: (1) NesT converges faster and requires much less training data to achieve good generalization on both ImageNet and small datasets like CIFAR; (2) when extending our key ideas to image generation, NesT leads to a strong decoder that is 8$\times$ faster than previous transformer-based generators; and (3) we show that decoupling the feature learning and abstraction processes via this nested hierarchy in our design enables constructing a novel method (named GradCAT) for visually interpreting the learned model. Source code is available https://github.com/google-research/nested-transformer.

Zizhao Zhang, Han Zhang, Long Zhao, Ting Chen, Sercan O. Arik, Tomas Pfister• 2021

Related benchmarks

TaskDatasetResultRank
Image ClassificationCIFAR-100 (test)
Accuracy82.56
3518
Image ClassificationImageNet (val)
Top-1 Acc83.8
1206
ClassificationImageNet-1K 1.0 (val)
Top-1 Accuracy (%)83.8
1155
Image ClassificationImageNet-1k (val)
Top-1 Accuracy86.2
840
Image ClassificationImageNet 1k (test)
Top-1 Accuracy83.8
798
Image ClassificationCIFAR-100 (val)--
661
Image ClassificationCIFAR10 (test)
Accuracy97.2
585
Image ClassificationImageNet-1k (val)
Top-1 Accuracy81.3
512
Image ClassificationCIFAR-10 (val)
Top-1 Accuracy97.2
329
Image ClassificationDomainNet
Accuracy (ClipArt)73.51
161
Showing 10 of 21 rows

Other info

Code

Follow for update