id
stringlengths 7
7
| images
sequencelengths 8
24
| text prompt
stringlengths 0
947
| video link
stringlengths 89
89
| visual quality
int64 1
4
| temporal consistency
int64 1
4
| dynamic degree
int64 1
4
| text-to-video alignment
int64 0
4
| factual consistency
int64 1
4
| conversations
listlengths 2
2
|
|---|---|---|---|---|---|---|---|---|---|
1002514 | [
"1002514_00.jpg",
"1002514_01.jpg",
"1002514_02.jpg",
"1002514_03.jpg",
"1002514_04.jpg",
"1002514_05.jpg",
"1002514_06.jpg",
"1002514_07.jpg",
"1002514_08.jpg",
"1002514_09.jpg",
"1002514_10.jpg",
"1002514_11.jpg",
"1002514_12.jpg",
"1002514_13.jpg",
"1002514_14.jpg",
"1002514_15.jpg"
] | villagers talking laughing dancing. Illustrative style. | https://huggingface.co/datasets/hexuan21/VideoFeedback-videos-mp4/blob/main/1/1002514.mp4 | 1 | 2 | 3 | 0 | 1 | [
{
"from": "human",
"value": "\nSuppose you are an expert in judging and evaluating the quality of AI-generated videos,\nplease watch the following frames of a given video and see the text prompt for generating the video,\nthen give scores from 5 different dimensions:\n(1) visual quality: the quality of the video in terms of clearness, resolution, brightness, and color\n(2) temporal consistency, both the consistency of objects or humans and the smoothness of motion or movements\n(3) dynamic degree, the degree of dynamic changes\n(4) text-to-video alignment, the alignment between the text prompt and the video content\n(5) factual consistency, the consistency of the video content with the common-sense and factual knowledge\n\nfor each dimension, output a float number from 1.0 to 4.0,\nthe higher the number is, the better the video performs in that sub-score, \nthe lowest 1.0 means Bad, the highest 4.0 means Perfect/Real (the video is like a real video)\nHere is an output example:\nvisual quality: 3.2\ntemporal consistency: 2.7\ndynamic degree: 4.0\ntext-to-video alignment: 2.3\nfactual consistency: 1.8\n\nFor this video, the text prompt is \"villagers talking laughing dancing. Illustrative style.\",\nall the frames of video are as follows:\n\n<image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> <image> "
},
{
"from": "gpt",
"value": "visual quality: 1\n temporal consistency: 2\n dynamic degree: 3\n text-to-video alignment: 0\n factual consistency: 1\n"
}
] |