Leopard-Idefic2

Paper | Github | Models-LLaVA | Models-Idefics2

Summaries

Text-rich images, where text serves as the central visual element guiding the overall understanding, are prevalent in real-world applications, such as presentation slides, scanned documents, and webpage snapshots. Tasks involving multiple text-rich images are especially challenging, as they require not only understanding the content of individual images but reasoning about inter-relationships and logical flows across multiple visual inputs. Despite the importance of these scenarios, current multimodal large language models (MLLMs) struggle to handle such tasks due to two key challenges: (1) the scarcity of high-quality instruction tuning datasets for text-rich multi-image scenarios, and (2) the difficulty in balancing image resolution with visual feature sequence length. To address these challenges, we propose Leopard, a MLLM designed specifically for handling vision-language tasks involving multiple text-rich images. First, we curated about one million high-quality multimodal instruction-tuning data, tailored to text-rich, multi-image scenarios. Second, we developed an adaptive high-resolution multi-image encoding module to dynamically optimize the allocation of visual sequence length based on the original aspect ratios and resolutions of the input images. Experiments across a wide range of benchmarks demonstrate our model's superior capabilities in text-rich, multi-image evaluations and competitive performance in general domain evaluations.

Architectures

For LEOPARD-Idefics2, we follow the architecture of Idefics2-8B which uses SigLIP-SO-400M as the visual en- coder but increases its image resolution to 980×980 to make the text legible. The features outputted by the visual encoder are compressed with a feature resampler into 64 tokens per image. Idefics2-8B adopts the Mistral-7B as the LM.

Citation

@article{jia2024leopard,
  title={LEOPARD: A Vision Language Model For Text-Rich Multi-Image Tasks},
  author={Jia, Mengzhao and Yu, Wenhao and Ma, Kaixin and Fang, Tianqing and Zhang, Zhihan and Ouyang, Siru and Zhang, Hongming and Jiang, Meng and Yu, Dong},
  journal={arXiv preprint arXiv:2410.01744},
  year={2024}
}