Pith. sign in

REVIEW 13 cited by

Vary: Scaling up the Vision Vocabulary for Large Vision-Language Models

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2312.06109 v1 pith:BGCJSIF4 submitted 2023-12-11 cs.CV

Vary: Scaling up the Vision Vocabulary for Large Vision-Language Models

classification cs.CV
keywords visionvocabularyvarylvlmsclipfeaturesfine-grainedlarge
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Modern Large Vision-Language Models (LVLMs) enjoy the same vision vocabulary -- CLIP, which can cover most common vision tasks. However, for some special vision task that needs dense and fine-grained vision perception, e.g., document-level OCR or chart understanding, especially in non-English scenarios, the CLIP-style vocabulary may encounter low efficiency in tokenizing the vision knowledge and even suffer out-of-vocabulary problem. Accordingly, we propose Vary, an efficient and effective method to scale up the vision vocabulary of LVLMs. The procedures of Vary are naturally divided into two folds: the generation and integration of a new vision vocabulary. In the first phase, we devise a vocabulary network along with a tiny decoder-only transformer to produce the desired vocabulary via autoregression. In the next, we scale up the vanilla vision vocabulary by merging the new one with the original one (CLIP), enabling the LVLMs can quickly garner new features. Compared to the popular BLIP-2, MiniGPT4, and LLaVA, Vary can maintain its vanilla capabilities while enjoying more excellent fine-grained perception and understanding ability. Specifically, Vary is competent in new document parsing features (OCR or markdown conversion) while achieving 78.2% ANLS in DocVQA and 36.2% in MMVet. Our code will be publicly available on the homepage.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 13 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. ShredBench: Evaluating the Semantic Reasoning Capabilities of Multimodal LLMs in Document Reconstruction

    cs.CV 2026-04 unverdicted novelty 7.0

    ShredBench shows state-of-the-art MLLMs perform well on intact documents but suffer sharp drops in restoration accuracy as fragmentation increases to 8-16 pieces, indicating insufficient cross-modal semantic reasoning...

  2. FinCriticalED: A Visual Benchmark for Financial Fact-Level OCR

    cs.CV 2025-11 unverdicted novelty 7.0

    FinCriticalED benchmark reveals that OCR and MLLM systems frequently fail to preserve critical financial facts such as numbers and monetary units even when lexical accuracy is high.

  3. Mixture of Cognitive Experts in Large Vision-Language Models

    cs.CV 2026-07 conditional novelty 6.0

    Routing CV experts into atomic evidence then Bloom-staged verbalization improves LVLM benchmarks and yields measurable query-conditioned reasoning traces.

  4. Improving Layout Representation Learning Across Inconsistently Annotated Datasets via Agentic Harmonization

    cs.CV 2026-04 unverdicted novelty 6.0

    VLM-based harmonization of inconsistent annotations across two document layout corpora raises detection F-score from 0.860 to 0.883 and table TEDS from 0.750 to 0.814 while tightening embedding clusters.

  5. Visual Funnel: Resolving Contextual Blindness in Multimodal Large Language Models

    cs.CV 2025-12 unverdicted novelty 6.0

    Visual Funnel resolves contextual blindness in MLLMs by constructing an entropy-scaled portfolio of hierarchically structured image crops that preserves both local detail and global context.

  6. General OCR Theory: Towards OCR-2.0 via a Unified End-to-end Model

    cs.CV 2024-09 unverdicted novelty 5.0

    GOT is a unified end-to-end model that treats all man-made optical signals as characters and handles multiple OCR tasks including formatted output and interactive region recognition via prompts.

  7. InternLM-XComposer-2.5: A Versatile Large Vision Language Model Supporting Long-Contextual Input and Output

    cs.CV 2024-07 conditional novelty 5.0

    InternLM-XComposer-2.5 is a 7B vision-language model supporting up to 96K context that reaches GPT-4V-level performance on image, video, and multi-turn tasks and adds LoRA-driven text-image composition capabilities.

  8. InternLM-XComposer2: Mastering Free-form Text-Image Composition and Comprehension in Vision-Language Large Model

    cs.CV 2024-01 unverdicted novelty 5.0

    InternLM-XComposer2 introduces Partial LoRA on InternLM2-7B to enable high-quality free-form text-image composition while matching or exceeding GPT-4V on select vision-language benchmarks.

  9. VideoLLaMA 3: Frontier Multimodal Foundation Models for Image and Video Understanding

    cs.CV 2025-01 unverdicted novelty 4.0

    VideoLLaMA3 uses a vision-centric training paradigm and token-reduction design to reach competitive results on image and video benchmarks.

  10. MinerU: An Open-Source Solution for Precise Document Content Extraction

    cs.CV 2024-09 conditional novelty 4.0

    MinerU delivers an open-source pipeline for high-precision document content extraction by integrating specialized models with tuned preprocessing and postprocessing rules.

  11. How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suites

    cs.CV 2024-04 unverdicted novelty 4.0

    InternVL 1.5 narrows the performance gap to proprietary multimodal models via a stronger transferable vision encoder, dynamic high-resolution tiling, and curated English-Chinese training data.

  12. MobileVLM V2: Faster and Stronger Baseline for Vision Language Model

    cs.CV 2024-02 unverdicted novelty 4.0

    MobileVLM V2 shows that 1.7B and 3B parameter vision-language models can reach or exceed the performance of 3B and 7B+ models on common VLM benchmarks via targeted design and data improvements.

  13. Document Parsing Unveiled: Techniques, Challenges, and Prospects for Structured Information Extraction

    cs.MM 2024-10 unverdicted novelty 3.0

    Survey proposing a taxonomy for document parsing into pipeline-based systems and VLM-driven unified models, reviewing components, metrics, benchmarks, and challenges.