| --- |
| license: apache-2.0 |
| task_categories: |
| - video-text-to-text |
| - visual-question-answering |
| - question-answering |
| language: |
| - en |
| tags: |
| - omni |
| - audio |
| - video |
| - audio-visual |
| - audio-centric |
| pretty_name: AVUTBenchmark (AV-Human split, repackaged for lmms-eval) |
| size_categories: |
| - 1K<n<10K |
| configs: |
| - config_name: default |
| data_files: |
| - split: test |
| path: av_human_data/test-* |
| dataset_info: |
| features: |
| - name: video_id |
| dtype: string |
| - name: question_id |
| dtype: string |
| - name: video_type |
| dtype: string |
| - name: task_type |
| dtype: string |
| - name: question |
| dtype: string |
| - name: options |
| dtype: string |
| - name: answer |
| dtype: string |
| - name: video_path |
| dtype: string |
| - name: subset_type |
| dtype: string |
| splits: |
| - name: test |
| num_bytes: 1050000 |
| num_examples: 1734 |
| --- |
| |
| # AVUTBenchmark |
|
|
| This repository provides the **AV-Human** split of the [**AVUT**](https://huggingface.co/datasets/tsinghua-ee/AVUTBenchmark) benchmark in a format compatible with [**lmms-eval**](https://github.com/EvolvingLMMs-Lab/lmms-eval). |
|
|
| The data is provided as a single parquet file containing only the QA annotations. Since raw videos are not included, please download them from the [original release](https://huggingface.co/datasets/tsinghua-ee/AVUTBenchmark) and match them with the QA annotations using `video_path`. |
|
|
| **Task configurations and evaluation scripts** are available in the SEATS repository: <https://github.com/xxayt/SEATS>. |
|
|
| ## Overview |
|
|
| AVUT is an audio-centric video understanding benchmark that evaluates multimodal LLMs on both audio content comprehension and audio-visual interaction. It introduces an answer permutation-based filtering mechanism to eliminate the "text shortcut problem" where answers can be inferred from question text alone. We use the **AV-Human** partition containing expert-annotated QA pairs across 18 audio-centric domains. |
|
|
| | Statistic | Value | |
| |------------------|----------------------------------------------------------------------| |
| | QA pairs | 1,734 (multiple choice, 4 options) | |
| | Videos | 691 videos, average ≈ 69 s | |
| | Video types | interview, music, education, sports, vlog, news, ... (18 domains) | |
| | Task types | Audio Character Matching, Audio OCR Matching, Audio Event Location, ...| |
| | Language | English | |
| | License | Apache-2.0 | |
|
|
| ## License |
|
|
| Released under the **Apache-2.0** license. The upstream |
| [tsinghua-ee/AVUTBenchmark](https://huggingface.co/datasets/tsinghua-ee/AVUTBenchmark) |
| release does not specify a license, so this repackaged QA split defaults to |
| Apache-2.0. |
|
|
