File size: 1,158 Bytes
e5d9030
1d43211
 
 
e5d9030
 
 
 
 
1d43211
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
---
title: ASID-Caption
emoji: 🦉
colorFrom: indigo
colorTo: gray
sdk: static
pinned: false
---

# ASID-Caption

We build **ASID-Caption**, a data-and-model suite for **fine-grained audiovisual video understanding**.

Our goal is to move beyond “one video → one generic caption” by providing **attribute-structured supervision** and **quality-verified annotations**, enabling models to produce **more complete, more controllable, and more temporally consistent** descriptions that cover both **visual content** and **audio cues**.

## What we release

- **ASID-1M**: a large-scale collection of **attribute-structured** audiovisual instructions with both *single-attribute* and *all-attributes* training formats.
- **ASID-Verify**: a scalable curation pipeline that generates, ensembles, verifies, and refines annotations to improve semantic and temporal consistency.
- **ASID-Captioner**: Qwen2.5-Omni-based audiovisual captioning models fine-tuned on ASID-1M.

## Research interests

- Video understanding & video captioning  
- Audio-visual learning  
- Multimodal LLMs / instruction tuning  
- Data curation, verification, and quality control