Shot2Story

### Clone with HTTP
```bash
# Get meta-data
git clone https://www.modelscope.cn/datasets/AI-ModelScope/Shot2Story.git
```

### Load dataset with sdk
```python
from modelscope import MsDataset
ds = MsDataset.load('AI-ModelScope/Shot2Story', split='train')
sample = next(iter(ds))
print(sample)
```

示例:
```shell
{'video': '7qwWoIWaMfs.7.mp4', 'image_id': '7qwWoIWaMfs.7.mp4', 'id': 0, 'whole_caption': "The video begins with a blonde woman in a black off-the-shoulder top engaging in a conversation with the camera. Simultaneously, a photograph is displayed on a screen to her right. The photograph features a woman with a double ponytail and a man, both dressed in black and wearing sunglasses, walking together. A person with green hair and sunglasses can be seen in the background of the photograph. The woman in the off-the-shoulder top is discussing Bella's recent comments about her relationship with Scott to Entertainment Tonight. The scene then shifts to an image of a blonde woman dressed in a white halter top and black leather pants. She is looking directly into the camera with a smile on her face. The audio reveals that she has commented on her relationship status, stating that they're just chilling and everything is cool. The video concludes with a shot of the long-haired woman wearing a necklace and a black jacket, looking straight at the camera. This is followed by a photo of a woman in a red knitted top, although her face is not visible. The audio provides further insight into the woman's relationship status, stating that she had just seen him and everything was still chill.", 'whole_ASR': 'That still seems to be unknown. Just last week, Bella had some interesting thing to say about her relationship with Scott to Entertainment Tonight, she said quote: were chilling were cool I literally just saw him, so its literally all chill she also mentioned.', 'nvid': '7qwWoIWaMfs.7.mp4', 'video_names': "['7qwWoIWaMfs.7_0_245.mp4', '7qwWoIWaMfs.7_246_335.mp4', '7qwWoIWaMfs.7_336_436.mp4']", 'audio_captions': '[\'According to the woman in the black off-the-shoulder top in the audio, Bella had some interesting things to say about her relationship with Scott to Entertainment Tonight just last week.\', "She said they\'re chilling, they\'re cool.", "She literally just saw him, so it\'s literally all chill."]', 'captions': "['The video shows a blonde woman in a black off-the-shoulder top talking to the camera, and on a screen to her right is a photograph of a woman in black with a double ponytail and a man in a black short-sleeved t-shirt and blue jeans, both wearing sunglasses, walking together. Behind them is a person with green hair and sunglasses.', 'The video shows an image of a blonde woman wearing a white halter top and black leather pants. She has a ponytail, looks straight into the camera and smiles.', 'The video shows a picture of a long-haired woman wearing a necklace and a black jacket, looking straight at the camera. The video then shows a photo of a woman wearing a red knitted top, her face is not shown.']", 'ASR': '[\' statuses, that still seems to be unknown. Just last week, Bella had some interesting things to say about her relationship with Scott to Entertainment Tonight.\', \' She said quote,\', " literally just saw him, so it\'s literally all chill."]', 'download_cmd': "youtube-dl -f 'bestvideo[ext=mp4][height<=360]+bestaudio/best[ext=mp4][height<=360]' -o 'train/7qwWoIWaMfs' https://www.youtube.com/watch?v=7qwWoIWaMfs"}

```

### Download video files
```python
import os

os.system(sample['download_cmd'])

```

### Reference
```shell
https://github.com/bytedance/Shot2Story/blob/master/DATA.md
```

### HTTP 克隆方式
bash
# 获取元数据
git clone https://www.modelscope.cn/datasets/AI-ModelScope/Shot2Story.git

### 使用SDK加载数据集
python
from modelscope import MsDataset
ds = MsDataset.load('AI-ModelScope/Shot2Story', split='train')
sample = next(iter(ds))
print(sample)

示例输出：
shell
{'视频': '7qwWoIWaMfs.7.mp4', '图像ID': '7qwWoIWaMfs.7.mp4', '样本ID': 0, '完整视频描述': "视频开篇，一位身着黑色露肩上衣的金发女性正对镜头交谈，同时其右侧屏幕展示了一张照片：照片中一位梳双马尾的女性与一名男性均身着黑色服饰并佩戴墨镜，正并肩行走，照片背景中可见一位戴墨镜的绿发人士。该女性正在《娱乐今夜》（Entertainment Tonight）节目中谈论贝拉近期针对她与斯科特的恋情发表的言论。场景随后切换至一位身着白色挂脖上衣与黑色皮裤的金发女性画面，她正对镜头微笑。音频中提及她已就恋情状态表态，称"我们只是在放松，一切都很好"。视频结尾，一位佩戴项链、身着黑色夹克的长发女性正对镜头，随后出现一张身着红色针织上衣的女性照片，其面部未露出。音频进一步补充了该女性的恋情状态：她刚见过对方，一切仍处于轻松状态。", '完整语音识别结果': "目前恋情状态仍不明确。就在上周，贝拉在《娱乐今夜》节目中谈及了她与斯科特的恋情，她说道："我们只是在放松，一切都很好，我真的刚见过他，所以一切都没问题。"她还补充道。", '视频唯一标识': '7qwWoIWaMfs.7.mp4', '视频分段文件名列表': "['7qwWoIWaMfs.7_0_245.mp4', '7qwWoIWaMfs.7_246_335.mp4', '7qwWoIWaMfs.7_336_436.mp4']", '音频字幕列表': '["据音频中的黑衣露肩上衣女性所述，贝拉就在上周于《娱乐今夜》节目中谈及了她与斯科特的恋情。", "她称二人正处于放松状态，一切都很好。", "她真的刚见过对方，所以一切都没问题。"]', '分段视频字幕列表': '["视频中，一位身着黑色露肩上衣的金发女性正对镜头交谈，其右侧屏幕上展示了一张照片：照片中一位梳双马尾、身着黑衣并佩戴墨镜的女性与一位身着黑色短袖T恤、蓝色牛仔裤并佩戴墨镜的男性并肩行走，二人身后背景处有一位戴墨镜的绿发人士。", "视频中展示了一位身着白色挂脖上衣、黑色皮裤的金发女性画面，她梳着马尾，正对镜头微笑。", "视频中展示了一位佩戴项链、身着黑色夹克的长发女性正对镜头的画面，随后出现一张身着红色针织上衣的女性照片，其面部未露出。"]', '分段语音识别结果列表': '["恋情状态目前仍不明确。就在上周，贝拉在《娱乐今夜》节目中谈及了她与斯科特的恋情。", " 她说道：", " 真的刚见过对方，所以一切都没问题。"]', '视频下载命令': "youtube-dl -f 'bestvideo[ext=mp4][height<=360]+bestaudio/best[ext=mp4][height<=360]' -o 'train/7qwWoIWaMfs' https://www.youtube.com/watch?v=7qwWoIWaMfs"}

### 下载视频文件
python
import os

os.system(sample['download_cmd'])

### 参考文献
shell
https://github.com/bytedance/Shot2Story/blob/master/DATA.md