While existing large vision-language multimodal models focus on whole image understanding, there is a prominent gap in achieving region-specific comprehension. Current approaches that use textual coordinates or spatial encodings often fail to provide a user-friendly interface for visual prompting. To address this challenge, we introduce a novel multimodal model capable of decoding arbitrary visual prompts. This allows users to intuitively mark images and interact with the model using natural cues like a "red bounding box" or "pointed arrow". Our simple design directly overlays visual markers onto the RGB image, eliminating the need for complex region encodings, yet achieves state-of-the-art performance on region-understanding tasks like Visual7W, PointQA, and Visual Commonsense Reasoning benchmark. Furthermore, we present ViP-Bench, a comprehensive benchmark to assess the capability of models in understanding visual prompts across multiple dimensions, enabling future research in this domain. Code, data, and model are publicly available.
ViP-LLaVA directly overlays the visual prompt with the original image, then feeds the image to the multimodal model. Our approach shows several benefits:
During Trianing, we use 8 diverse visual prompts, including mask contour, ellipse, bounding box, triangle, scribble, point, arrow, and mask. Note that the prompts not only have diverse shapes, but they also have diverse colors, transparency values, widths, scales, and directions.
ViP-Bench is the first zero-shot region-level benchmarks that comprehensively evaluate the capability of multimodal models in understanding visual prompts. ViP-Bench is composed of 303 samples, evaluating 6 capbilities including recognition, OCR, knowledge, math, object relationship reason- ing, and language generation. ViP-Bench have two formats: (1) Bounding box format, and (2) arbitrary visual prompts annotated by human.
Here we should some cases where GPT-4V fails to recognize the visual prompts while ViP-LLaVA succeeds. However, in most cases, GPT-4V shows robust and strong performance in visual prompt understanding.
@inproceedings{cai2023vipllava,
author = {Cai, Mu and Liu, Haotian and Mustikovela, Siva Karthik and Meyer, Gregory P. and Chai, Yuning and Park, Dennis and Lee, Yong Jae},
title = {Making Large Multimodal Models Understand Arbitrary Visual Prompts},
booktitle = {IEEE Conference on Computer Vision and Pattern Recognition},
year = {2024}
}
This website is adapted from Nerfies, licensed under a Creative Commons Attribution-ShareAlike 4.0 International License. We thank the LLaMA team for giving us access to their models, and open-source projects, including Alpaca and Vicuna.
Usage and License Notices: The data, code and checkpoint is intended and licensed for research use only. They are also restricted to uses that follow the license agreement of CLIP, LLaMA, Vicuna and GPT-4. The dataset is CC BY NC 4.0 (allowing only non-commercial use) and models trained using the dataset should not be used outside of research purposes.
Related Links: [LLaVA] [Insutrction Tuning with GPT-4]