The convergence of text, visual, and audio data is a key step towards human-like artificial intelligence, however the current Vision-Language-Speech landscape is dominated by encoder-only models which lack generative abilities. We propose closing this gap with i-Code V2, the first model capable of generating natural language from any combination of Vision, Language, and Speech data. i-Code V2 is an integrative system that leverages state-of-the-art single-modality encoders, combining their outputs with a new modality-fusing encoder in order to flexibly project combinations of modalities into a shared representational space. Next, language tokens are generated from these representations via an autoregressive decoder. The whole framework is pretrained end-to-end on a large collection of dual- and single-modality datasets using a novel text completion objective that can be generalized across arbitrary combinations of modalities. i-Code V2 matches or outperforms state-of-the-art single- and dual-modality baselines on 7 multimodal tasks, demonstrating the power of generative multimodal pretraining across a diversity of tasks and signals.
@article{arxiv.2305.12311,
title = {i-Code V2: An Autoregressive Generation Framework over Vision, Language, and Speech Data},
author = {Ziyi Yang and Mahmoud Khademi and Yichong Xu and Reid Pryzant and Yuwei Fang and Chenguang Zhu and Dongdong Chen and Yao Qian and Mei Gao and Yi-Ling Chen and Robert Gmyr and Naoyuki Kanda and Noel Codella and Bin Xiao and Yu Shi and Lu Yuan and Takuya Yoshioka and Michael Zeng and Xuedong Huang},
journal= {arXiv preprint arXiv:2305.12311},
year = {2023}
}