Datasets:
File size: 5,252 Bytes
a7a4d45 d05d50b a7a4d45 d05d50b a7a4d45 | 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 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 | ---
task_categories:
- automatic-speech-recognition
language:
- en
- zh
tags:
- multimodal
- avsr
- video-conferencing
- lombard-effect
---
# π₯ MLD-VC: Multimodal Dataset for Video Conferencing
> **When AVSR Meets Video Conferencing: Dataset, Degradation, and the Hidden Mechanism Behind Performance Collapse (CVPR 2026)**
> π [[Paper\]](https://arxiv.org/abs/2603.22915) | π€ [[Hugging Face Dataset\]](https://huggingface.co/datasets/nccm2p2/MLD-VC)
------
## π Overview
**MLD-VC** is the **first multimodal dataset specifically designed for Audio-Visual Speech Recognition (AVSR) in real-world video conferencing (VC) scenarios**.
Unlike traditional AVSR datasets collected in controlled offline environments, MLD-VC explicitly models two critical factors in VC:
- **Transmission Distortions** (compression, speech enhancement, etc.)
- **Human Hyper-expression** (e.g., Lombard effect)
### π Key Features
- π€ **31 speakers**, 22.79 hours of recordings
- π **4 mainstream VC platforms**
- π£οΈ **Bilingual**: English & Chinese
- π§ **Lombard effect simulation** via noise conditions
- π₯ Multimodal data:
- Video
- Audio
- Facial landmarks
- text
------
## π¨ Motivation
Existing AVSR systems show **severe performance degradation in video conferencing**, due to:
- Distribution shift caused by **speech enhancement algorithms**
- Behavioral changes such as **hyper-expression**
MLD-VC is designed to **bridge the gap between offline datasets and real-world VC deployment**.\
------
## π Dataset Structure
The dataset is organized into three aligned modalities:
```
MLD-VC/
βββ video/
βββ audio/
βββ landmarks/
```
Each modality follows the **same hierarchical structure**:
```
<modality>/
βββ Online / Offline
βββ speaker_id
βββ platform
βββ sentence_id
βββ clean / 40db / 60db / 80db
```
### π Example
```
video/
βββ Online/
βββ speaker_03/
βββ Zoom/
βββ sentence_012/
βββ clean/
βββ 40db/
βββ 60db/
βββ 80db/
```
------
## π§ Data Description
### 1. Online vs Offline
- **Offline**:
- Direct recording (no transmission)
- Contains hyper-expression (via noise)
- **Online**:
- Recorded after transmission through VC platforms
- Includes:
- Compression
- Speech enhancement
- Network effects
------
### 2. Noise Levels (Lombard Effect)
Each sentence is recorded under 4 noise conditions:
| Condition | Description |
| --------- | -------------- |
| clean | No noise |
| 40dB | Mild noise |
| 60dB | Moderate noise |
| 80dB | Strong noise |
These simulate **Lombard effect intensity**, inducing hyper-expression.
------
### 3. Platforms
The dataset includes recordings from multiple VC platforms (e.g.):
- Zoom
- Tencent Meeting
- Lark
- DingTalk
------
## β οΈ Important Notes
### π Recording Protocol Differences
- In **Offline subset**:
- **Speakers 2β8**:
- Recorded on **a single device**, repeated across 4 platforms
- Other speakers:
- **DD platform only**, but actually recorded using **4 different devices simultaneously**
π This leads to:
- Platform variation β always device variation
- Be careful in **cross-platform generalization experiments**
------
### β Removed Speakers
- **Speaker 0 and 1 have been removed**
- Due to poor recording quality
------
### π Data Consistency
- All three modalities (`video`, `audio`, `landmarks`) are:
- **Strictly aligned**
- Share identical folder structure
- Can be indexed jointly
------
## π¬ Recommended Use Cases
MLD-VC is suitable for:
### β AVSR Robustness
- Evaluate performance under real VC conditions
### β Cross-domain Generalization
- Train on Offline β Test on Online
### β Multimodal Learning
- Audio-visual fusion
- Landmark-based modeling
### β Distribution Shift Analysis
- Study impact of:
- Speech enhancement
- Lombard effect
------
## π Key Findings (from the paper)
- AVSR models suffer **massive degradation in VC**
- **Speech enhancement** is the main cause of audio distribution shift
- **Lombard effect β VC distortion (in feature space)**
- Landmark-based features are **more stable than image features**
- Fine-tuning on MLD-VC reduces CER by **17.5%**
------
## π Citation
If you find this dataset useful, please cite:
```bibtex
@inproceedings{huang2026mldvc,
title={When AVSR Meets Video Conferencing: Dataset, Degradation, and the Hidden Mechanism Behind Performance Collapse},
author={Huang, Yihuan and Xue, Jun and Liu, Jiajun and Li, Daixian and Zhang, Tong and Yi, Zhuolin and Ren, Yanzhen and Li, Kai},
booktitle={CVPR},
year={2026}
}
```
------
## π Acknowledgements
This work is supported by:
- National Natural Science Foundation of China
- DiDi Chuxing Group
------
## π¬ Contact
If you have questions, feel free to contact:
- **Yihuan Huang**: [yihuanhuang@whu.edu.cn](mailto:yihuanhuang@whu.edu.cn)
------
## β Star This Repo
If you find MLD-VC helpful, please consider giving a β! |