Learning by Correction: Efficient Tuning Task for Zero-Shot Generative Vision-Language Reasoning

arXiv:2404.00909v1 Announce Type: new
Abstract: Generative vision-language models (VLMs) have shown impressive performance in zero-shot vision-language tasks like image captioning and visual question answering. However, improving their zero-shot reasoning typically requires second-stage instruction tuning, which relies heavily on human-labeled or large language model-generated annotation, incurring high labeling costs. To tackle this challenge, we introduce Image-Conditioned Caption Correction (ICCC), a novel pre-training task designed to enhance VLMs' zero-shot performance without the need for labeled task-aware data. The ICCC task compels VLMs …

abstract annotation arxiv captioning costs cs.cv generated generative however human image improving labeling language language model language models large language large language model performance question question answering reasoning stage tasks type vision vision-language models visual vlms zero-shot