Multimodal Conditionality for Natural Language Generation
원문 초록
Large scale pretrained language models have demonstrated state-of-the-art performance in language understanding tasks. Their application has recently expanded into multimodality learning, leading to improved representations combining vision and language. However, progress in adapting language models towards conditional Natural Language Generation (NLG) has been limited to a single modality, generally text. We propose MAnTiS, Multimodal Adaptation for Text Synthesis, a general approach for multimodal conditionality in transformer-based NLG models. In this method, we pass inputs from each modality through modality-specific encoders, project to textual token space, and finally join to form a conditionality prefix. We fine-tune the pretrained language model and encoders with the conditionality prefix guiding the generation. We apply MAnTiS to the task of product description generation, conditioning a network on both product images and titles to generate descriptive text. We demonstrate that MAnTiS outperforms strong baseline approaches on standard NLG scoring metrics. Furthermore, qualitative assessments demonstrate that MAnTiS can generate human quality descriptions consistent with given multimodal inputs.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
Making Language Model a Hierarchical Classifier and Generator
Describe Anything Model for Visual Question Answering on Text-rich Images
Assay2Mol: large language model-based drug design using BioAssay context
CoLMbo: Speaker Language Model for Descriptive Profiling
TagRouter: Learning Route to LLMs through Tags for Open-Domain Text Generation Tasks
Towards Secure and Private Language Models for Nuclear Power Plants
토론
이 논문과 연결된 포럼 글이 아직 없습니다.