Harmonizing Visual Text Comprehension and Generation
This addresses the problem of inefficient multimodal generation for AI researchers and practitioners by offering a more unified approach, though it appears incremental as it builds on existing LoRA techniques.
The paper tackles the problem of performance degradation when simultaneously generating images and texts due to modality inconsistency, proposing Slide-LoRA to harmonize vision and language generation in a single model. The result is comparable performance to modality-specific fine-tuning with only a 2% parameter increase, achieving average improvements of 2.5% in visual text comprehension and 4.0% in visual text generation tasks.
In this work, we present TextHarmony, a unified and versatile multimodal generative model proficient in comprehending and generating visual text. Simultaneously generating images and texts typically results in performance degradation due to the inherent inconsistency between vision and language modalities. To overcome this challenge, existing approaches resort to modality-specific data for supervised fine-tuning, necessitating distinct model instances. We propose Slide-LoRA, which dynamically aggregates modality-specific and modality-agnostic LoRA experts, partially decoupling the multimodal generation space. Slide-LoRA harmonizes the generation of vision and language within a singular model instance, thereby facilitating a more unified generative process. Additionally, we develop a high-quality image caption dataset, DetailedTextCaps-100K, synthesized with a sophisticated closed-source MLLM to enhance visual text generation capabilities further. Comprehensive experiments across various benchmarks demonstrate the effectiveness of the proposed approach. Empowered by Slide-LoRA, TextHarmony achieves comparable performance to modality-specific fine-tuning results with only a 2% increase in parameters and shows an average improvement of 2.5% in visual text comprehension tasks and 4.0% in visual text generation tasks. Our work delineates the viability of an integrated approach to multimodal generation within the visual text domain, setting a foundation for subsequent inquiries. Code is available at https://github.com/bytedance/TextHarmony.