DialectGen: Benchmarking and Improving Dialect Robustness in Multimodal Generation
2510.14949v1
cs.CL, cs.CV, cs.LG
2025-10-18
Авторы:
Yu Zhou, Sohyun An, Haikang Deng, Da Yin, Clark Peng, Cho-Jui Hsieh, Kai-Wei Chang, Nanyun Peng
Abstract
Contact languages like English exhibit rich regional variations in the form
of dialects, which are often used by dialect speakers interacting with
generative models. However, can multimodal generative models effectively
produce content given dialectal textual input? In this work, we study this
question by constructing a new large-scale benchmark spanning six common
English dialects. We work with dialect speakers to collect and verify over 4200
unique prompts and evaluate on 17 image and video generative models. Our
automatic and human evaluation results show that current state-of-the-art
multimodal generative models exhibit 32.26% to 48.17% performance degradation
when a single dialect word is used in the prompt. Common mitigation methods
such as fine-tuning and prompt rewriting can only improve dialect performance
by small margins (< 7%), while potentially incurring significant performance
degradation in Standard American English (SAE). To this end, we design a
general encoder-based mitigation strategy for multimodal generative models. Our
method teaches the model to recognize new dialect features while preserving SAE
performance. Experiments on models such as Stable Diffusion 1.5 show that our
method is able to simultaneously raise performance on five dialects to be on
par with SAE (+34.4%), while incurring near zero cost to SAE performance.
Ссылки и действия
Дополнительные ресурсы: