Gamified crowd-sourcing of high-quality data for visual fine-tuning
Keywords: Large Multimodal Models, Visual Question Answering, Visual Instruction Tuning, Gamification, Supervised Learning, Data Generation
TL;DR: GAP is a novel framework for crowdsourcing high-quality visual instruction data by gamifying the process. It encourages users to create challenging questions, improving model accuracy and performance across various benchmarks and models
Abstract: This paper introduces gamified adversarial prompting (GAP), a framework that
crowd-sources high-quality data for visual instruction tuning of large multimodal
models. GAP transforms the data collection process into an engaging game, in-
centivizing players to provide fine-grained, challenging questions and answers
that target gaps in the model’s knowledge. Our contributions include (1) an ap-
proach to capture question-answer pairs from humans that directly address weak-
nesses in a model’s knowledge, (2) a method for evaluating and rewarding players
that successfully incentivizes them to provide high-quality submissions, and (3) a
scalable, gamified platform that succeeds in collecting this data from over 50,000
participants in just a few weeks. Our implementation of GAP has significantly im-
proved the accuracy of a small multimodal model, namely MiniCPM-Llama3-V-
2.5-8B, increasing its GPT score from 0.147 to 0.477 on our dataset, approaching
the benchmark set by the much larger GPT-4V. Moreover, we demonstrate that
the data generated using MiniCPM-Llama3-V-2.5-8B also enhances its perfor-
mance across other benchmarks, and exhibits cross-model benefits. Specifically,
the same data improves the performance of QWEN2-VL-2B and QWEN2-VL-7B
on the same multiple benchmarks.
Primary Area: applications to computer vision, audio, language, and other modalities
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.
Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.
Reciprocal Reviewing: I understand the reciprocal reviewing requirement as described on https://iclr.cc/Conferences/2025/CallForPapers. If none of the authors are registered as a reviewer, it may result in a desk rejection at the discretion of the program chairs. To request an exception, please complete this form at https://forms.gle/Huojr6VjkFxiQsUp6.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Submission Number: 11677
Loading