Exploiting Multiple Sequence Lengths in Fast End to End Training for Image Captioning

TL;DR

This paper presents ExpansionNet v2, a novel image captioning architecture that leverages multiple sequence lengths for improved learning and efficiency, achieving state-of-the-art results and faster training times.

Contribution

Introduction of the Expansion mechanism and ExpansionNet v2 architecture for more effective and faster end-to-end image captioning training.

Findings

Achieved top scores on MS COCO 2014 challenge

End-to-end training up to 2.8 times faster

State-of-the-art CIDErD scores on multiple datasets

Abstract

We introduce a method called the Expansion mechanism that processes the input unconstrained by the number of elements in the sequence. By doing so, the model can learn more effectively compared to traditional attention-based approaches. To support this claim, we design a novel architecture ExpansionNet v2 that achieved strong results on the MS COCO 2014 Image Captioning challenge and the State of the Art in its respective category, with a score of 143.7 CIDErD in the offline test split, 140.8 CIDErD in the online evaluation server and 72.9 AllCIDEr on the nocaps validation set. Additionally, we introduce an End to End training algorithm up to 2.8 times faster than established alternatives. Source code available at: https://github.com/jchenghu/ExpansionNet_v2