以结构化 JSON 格式提取视频中的章节
|
在对 Cloud 控制台使用不同参数值时，或通过直接调用 Vertex AI API，您可以直接查询模型，并测试返回的结果。
|
|||||||||||
提示自由格式
通过将视频内容分组为章节并提供每个章节的摘要，对视频内容进行章节分类。请仅拍摄重要活动和亮点。如果您不确定任何信息，请不要编造它。返回带有如下键的 JSON 格式的结果：“timecode”“chapterSummary”
响应
[
{
"timecode": "0:00-0:19",
"chapterSummary": "A panoramic view of Rio de Janeiro, highlighting the iconic Christ the Redeemer statue, Copacabana Beach, and the vibrant city life."
},
{
"timecode": "0:19-0:28",
"chapterSummary": "Introduction to the favelas of Rio, a significant part of the city often depicted negatively in the media."
},
{
"timecode": "0:28-0:37",
"chapterSummary": "Challenging the stereotypical portrayal of favelas, showcasing the everyday lives and resilience of its inhabitants."
},
{
"timecode": "0:37-0:49",
"chapterSummary": "Google's initiative to map the favelas, emphasizing the importance of recognizing their identity and address."
},
{
"timecode": "0:49-0:57",
"chapterSummary": "An invitation to explore the favelas through Google's 360° virtual tour."
}
]
|
|
||||||||||