Update README.md
Browse files
README.md
CHANGED
|
@@ -32,7 +32,16 @@ For the first two benchmarks, we take Spearman corrleation between model's outpu
|
|
| 32 |
averaged among all the evaluation aspects as indicator.
|
| 33 |
For GenAI-Bench and VBench, which include human preference data among two or more videos,
|
| 34 |
we employ the model's output to predict preferences and use pairwise accuracy as the performance indicator.
|
| 35 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 36 |
|:-----------------:|:---------------:|:--------------:|:-----------:|:-----------:|:----------:|
|
| 37 |
| MantisScore (reg) | **278.3** | 75.7 | **51.1** | **78.5** | **73.0** |
|
| 38 |
| MantisScore (gen) | 222.4 | **77.1** | 27.6 | 59.0 | 58.7 |
|
|
|
|
| 32 |
averaged among all the evaluation aspects as indicator.
|
| 33 |
For GenAI-Bench and VBench, which include human preference data among two or more videos,
|
| 34 |
we employ the model's output to predict preferences and use pairwise accuracy as the performance indicator.
|
| 35 |
+
|
| 36 |
+
Moreover, we use [MantisScore](https://huggingface.co/TIGER-Lab/MantisScore) trained on VideoFeedback dataset
|
| 37 |
+
for VideoFeedback-test set, while for other three benchmarks, we use
|
| 38 |
+
[MantisScore-anno-only](https://huggingface.co/TIGER-Lab/MantisScore-anno-only) variant trained on VideoFeedback dataset
|
| 39 |
+
with real videos excluded.
|
| 40 |
+
|
| 41 |
+
The evaluation results are shown below:
|
| 42 |
+
|
| 43 |
+
|
| 44 |
+
| metric | Final Sum Score | VideoFeedback-test | EvalCrafter | GenAI-Bench | VBench |
|
| 45 |
|:-----------------:|:---------------:|:--------------:|:-----------:|:-----------:|:----------:|
|
| 46 |
| MantisScore (reg) | **278.3** | 75.7 | **51.1** | **78.5** | **73.0** |
|
| 47 |
| MantisScore (gen) | 222.4 | **77.1** | 27.6 | 59.0 | 58.7 |
|