Ground-V: Teaching VLMs to Ground Complex Instructions in Pixels

• URL: http://arxiv.org/abs/2505.13788v1
• Date: Tue, 20 May 2025 00:37:19 GMT
• Title: Ground-V: Teaching VLMs to Ground Complex Instructions in Pixels
• Authors: Yongshuo Zong, Qin Zhang, Dongsheng An, Zhihua Li, Xiang Xu, Linghan Xu, Zhuowen Tu, Yifan Xing, Onkar Dabeer,
• Abstract summary: We address five critical real-world challenges in text-instruction-based grounding.<n>Our approach generates high-quality instruction-response pairs linked to existing pixel-level annotations.<n>Experiment results show that models trained on Ground-V exhibit substantial improvements across diverse grounding tasks.
• Score: 30.722073025794025
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: This work presents a simple yet effective workflow for automatically scaling instruction-following data to elicit pixel-level grounding capabilities of VLMs under complex instructions. In particular, we address five critical real-world challenges in text-instruction-based grounding: hallucinated references, multi-object scenarios, reasoning, multi-granularity, and part-level references. By leveraging knowledge distillation from a pre-trained teacher model, our approach generates high-quality instruction-response pairs linked to existing pixel-level annotations, minimizing the need for costly human annotation. The resulting dataset, Ground-V, captures rich object localization knowledge and nuanced pixel-level referring expressions. Experiment results show that models trained on Ground-V exhibit substantial improvements across diverse grounding tasks. Specifically, incorporating Ground-V during training directly achieves an average accuracy boost of 4.4% for LISA and a 7.9% for PSALM across six benchmarks on the gIoU metric. It also sets new state-of-the-art results on standard benchmarks such as RefCOCO/+/g. Notably, on gRefCOCO, we achieve an N-Acc of 83.3%, exceeding the previous state-of-the-art by more than 20%.

Related papers

• GroundingSuite: Measuring Complex Multi-Granular Pixel Grounding [39.967352995143855]
  GroundingSuite aims to bridge the gap between vision and language modalities.<n>It comprises: (1) an automated data annotation framework leveraging multiple Vision-Language Model (VLM) agents; (2) a large-scale training dataset encompassing 9.56 million diverse referring expressions and their corresponding segmentations; and (3) a meticulously curated evaluation benchmark consisting of 3,800 images.
  arXiv  Detail & Related papers  (2025-03-13T17:43:10Z)
• SURDS: Benchmarking Spatial Understanding and Reasoning in Driving Scenarios with Vision Language Models [15.50826328938879]
  We introduce SURDS, a benchmark designed to evaluate the spatial reasoning capabilities of vision language models (VLMs)<n>Built on the nuScenes dataset, SURDS comprises 41,080 vision-question-answer training instances and 9,250 evaluation samples.<n>We propose a reinforcement learning-based alignment scheme leveraging spatially grounded reward signals.
  arXiv  Detail & Related papers  (2024-11-20T08:14:01Z)
• RS-GPT4V: A Unified Multimodal Instruction-Following Dataset for Remote Sensing Image Understanding [4.266920365127677]
  Under the new LaGD paradigm, the old datasets are no longer suitable for fire-new tasks. We designed a high-quality, diversified, and unified multimodal instruction-following dataset for RSI understanding. The empirical results show that the fine-tuned MLLMs by RS-GPT4V can describe fine-grained information.
  arXiv  Detail & Related papers  (2024-06-18T10:34:28Z)
• Less is More: High-value Data Selection for Visual Instruction Tuning [127.38740043393527]
  We propose a high-value data selection approach TIVE, to eliminate redundancy within the visual instruction data and reduce the training cost. Our approach using only about 15% data can achieve comparable average performance to the full-data fine-tuned model across eight benchmarks.
  arXiv  Detail & Related papers  (2024-03-14T16:47:25Z)
• GroundVLP: Harnessing Zero-shot Visual Grounding from Vision-Language Pre-training and Open-Vocabulary Object Detection [24.48128633414131]
  We propose a zero-shot method that harnesses visual grounding ability from existing models trained from image-text pairs and pure object detection data. We demonstrate that the proposed method significantly outperforms other zero-shot methods on RefCOCO/+/g datasets.
  arXiv  Detail & Related papers  (2023-12-22T20:14:55Z)
• Silkie: Preference Distillation for Large Visual Language Models [56.10697821410489]
  This paper explores preference distillation for large vision language models (LVLMs) We first build a vision-language feedback dataset utilizing AI annotation. We adopt GPT-4V to assess the generated outputs regarding helpfulness, visual faithfulness, and ethical considerations. The resulting model Silkie, achieves 6.9% and 9.5% relative improvement on the MME benchmark regarding the perception and cognition capabilities.
  arXiv  Detail & Related papers  (2023-12-17T09:44:27Z)
• Pink: Unveiling the Power of Referential Comprehension for Multi-modal LLMs [49.88461345825586]
  This paper proposes a new framework to enhance the fine-grained image understanding abilities of MLLMs. We present a new method for constructing the instruction tuning dataset at a low cost by leveraging annotations in existing datasets. We show that our model exhibits a 5.2% accuracy improvement over Qwen-VL and surpasses the accuracy of Kosmos-2 by 24.7%.
  arXiv  Detail & Related papers  (2023-10-01T05:53:15Z)
• AnomalyGPT: Detecting Industrial Anomalies Using Large Vision-Language Models [30.723122000372538]
  AnomalyGPT is a novel IAD approach based on Large Vision-Language Models (LVLM) We generate training data by simulating anomalous images and producing corresponding textual descriptions for each image. AnomalyGPT achieves the state-of-the-art performance with an accuracy of 86.1%, an image-level AUC of 94.1%, and a pixel-level AUC of 95.3% on the MVTec-AD dataset.
  arXiv  Detail & Related papers  (2023-08-29T15:02:53Z)
• MOCA: Self-supervised Representation Learning by Predicting Masked Online Codebook Assignments [72.6405488990753]
  Self-supervised learning can be used for mitigating the greedy needs of Vision Transformer networks. We propose a single-stage and standalone method, MOCA, which unifies both desired properties. We achieve new state-of-the-art results on low-shot settings and strong experimental results in various evaluation protocols.
  arXiv  Detail & Related papers  (2023-07-18T15:46:20Z)
• Improving Visual Grounding by Encouraging Consistent Gradient-based Explanations [58.442103936918805]
  We show that Attention Mask Consistency produces superior visual grounding results than previous methods. AMC is effective, easy to implement, and is general as it can be adopted by any vision-language model.
  arXiv  Detail & Related papers  (2022-06-30T17:55:12Z)
• Propagate Yourself: Exploring Pixel-Level Consistency for Unsupervised Visual Representation Learning [60.75687261314962]
  We introduce pixel-level pretext tasks for learning dense feature representations. A pixel-to-propagation consistency task produces better results than state-of-the-art approaches. Results demonstrate the strong potential of defining pretext tasks at the pixel level.
  arXiv  Detail & Related papers  (2020-11-19T18:59:45Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.