Documentation changes #450
Conversation
Coverage reportNote Coverage evolution disabled because this PR targets a different branch The coverage rate is None of the new lines are part of the tested code. Therefore, there is no coverage data about them. |
| @@ -173,6 +180,14 @@ The following table contains the list of supported datasets. | |||
| - multiple | |||
| - Any `Hugging Face dataset <https://huggingface.co/datasets>`__ can be used. Just prepend the prefix ``hfd-``, e.g., ``hfd-rotten_tomatoes``. Select input and target columns via ``config_space``, e.g., add ``"input_column": "text", "target_column": "label"`` for the `rotten_tomatoes <https://huggingface.co/datasets/rotten_tomatoes>`__ example. | |||
| - Please refer to `the official documentation <https://huggingface.co/datasets>`__. | |||
| * - CDDB | |||
| - Image Classification | |||
| - 2 classes from 5 domains of generated images using various techniques. Numbers vary across domains. | |||
There was a problem hiding this comment.
can we make domains explicit?
"2 classes from 5 domains of generated images using various techniques" this description wouldn't be clear to me. how about this:
- Image Classification: deepfake detection
- 2 classes, 5 domains, each representing an image generation technique: ...
| - 2 classes from 5 domains of generated images using various techniques. Numbers vary across domains. | ||
| - Li, Chuqiao, et al. A continual deepfake detection benchmark: Dataset, methods, and essentials. IEEE/CVF Winter Conference on Applications of Computer Vision. 2023. | ||
| * - CORe50 | ||
| - Object Recognition - image classfication |
There was a problem hiding this comment.
Can we stick to the format used earlier, i.e., Image Classification?
| - Image Classification | ||
| - 2 classes from 5 domains of generated images using various techniques. Numbers vary across domains. | ||
| - Li, Chuqiao, et al. A continual deepfake detection benchmark: Dataset, methods, and essentials. IEEE/CVF Winter Conference on Applications of Computer Vision. 2023. | ||
| * - CORe50 |
There was a problem hiding this comment.
This refers to the name we use to load the data as part of benchmarking. Right now, this should be Core50.
| - Li, Chuqiao, et al. A continual deepfake detection benchmark: Dataset, methods, and essentials. IEEE/CVF Winter Conference on Applications of Computer Vision. 2023. | ||
| * - CORe50 | ||
| - Object Recognition - image classfication | ||
| - 10 classes of 7 training domains, 3 other domains used for evaluation |
There was a problem hiding this comment.
There are 8 domains. The 3 other domains is irrelevant for evaluation. Rather mention that there is only 1 test dataset which will be used as test set for each task. Same comment as before "10 classes of 7 training domains". What does that mean? Rather something like "50 classes, 8 domains (0-7)." Note, there are 50 classes, not 10.
Adds CDDB, Core50, L2P documentation on the benchmarks listing page.
By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.