SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2412.13871 · 2024

LLaVA-UHD v2: an MLLM Integrating High-Resolution Semantic Pyramid via Hierarchical Window Transformer

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 11 functions out of this paper's own repositories and ran 7 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
thunlp/llava-uhd canonical 7 of 11
FunctionStatusWhere it lives
extract_answer Ran thunlp/llava-uhd/VLMEvalKit/vlmeval/vlm/llava_qwen2_uhd_v3.py
code served (permissive licence) · get_code("a76fe5da7ade58fb")
extract_choice Ran thunlp/llava-uhd/VLMEvalKit/vlmeval/vlm/llava_qwen2_uhd_v3.py
code served (permissive licence) · get_code("46ca4942d73607af")
extract_patches Ran thunlp/llava-uhd/llava/mm_utils.py
code served (permissive licence) · get_code("b0a851cae92754ff")
extract_yes_no Ran thunlp/llava-uhd/VLMEvalKit/vlmeval/vlm/llava_qwen2_uhd_v3.py
code served (permissive licence) · get_code("fb3f8b5f16ed11cb")
get_2d_sincos_pos_embed Ran thunlp/llava-uhd/llava/model/multimodal_projector/uhd_v1_resampler.py
code served (permissive licence) · get_code("77e8a3ac46f3afec")
get_abs_pos Ran thunlp/llava-uhd/llava/model/multimodal_projector/uhd_v1_resampler.py
code served (permissive licence) · get_code("17e90d5853a48fa7")
unpad_image Ran thunlp/llava-uhd/llava/model/llava_arch.py
code served (permissive licence) · get_code("7606525af238fb64")
auto_pad_images Not yet run thunlp/llava-uhd/llava/mm_utils.py
code served (permissive licence) · get_code("3c8823948724e269")
get_2d_sincos_pos_embed_from_grid Not yet run thunlp/llava-uhd/llava/model/multimodal_projector/uhd_v1_resampler.py
code served (permissive licence) · get_code("2884deadbb637c3e")
resize_and_center_crop Not yet run thunlp/llava-uhd/llava/mm_utils.py
code served (permissive licence) · get_code("c80358dff96c25ce")
safe_load_tokenizer Not yet run thunlp/llava-uhd/llava/conversation.py
code served (permissive licence) · get_code("b20705b0dfd77ffc")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Vision transformers (ViTs) are widely employed in multimodal large language models (MLLMs) for visual encoding. However, they exhibit inferior performance on tasks regarding fine-grained visual perception. We attribute this to the limitations of ViTs in capturing diverse multi-modal visual levels, such as low-level details. To address this issue, we present LLaVA-UHD v2, an MLLM with advanced perception abilities by introducing a well-designed vision-language projector, the Hierarchical window (Hiwin) transformer. Hiwin transformer enhances MLLM's ability to capture diverse multi-modal visual granularities, by incorporating our constructed high-resolution semantic pyramid. Specifically, Hiwin transformer comprises two key modules: (i) a visual detail injection module, which progressively injects low-level visual details into high-level language-aligned semantics features, thereby forming an inverse semantic pyramid (ISP), and (ii) a hierarchical window attention module, which leverages cross-scale windows to condense multi-level semantics from the ISP. Extensive experiments show that LLaVA-UHD v2 outperforms compared MLLMs on a wide range of benchmarks. Notably, our design achieves an average boost of 3.7% across 14 benchmarks compared with the baseline method, 9.3% on DocVQA for instance. All the data and code will be publicly available to facilitate future research.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2412.13871")
get_code_for_paper("2412.13871")
have("2412.13871")

Connect an agent — have() is free.