-
微博内容精选
-
I have followed your work with Flowtron and it produces audio really nicely. I am using the same dataset of 1.5 hrs of a single speaker but I have trouble getting some output that is not as muffled he…
-
-
## Summary
The rolls on the substats of equipments on the enhancement page will shift left upon enhancement of equipment. This causes the OCR to read the orange text as well such as in the screenshot…
joszx updated
2 years ago
-
Hello, this is a great work. But when I use weights provided in the link [Baidu Driver](https://pan.baidu.com/s/1GRUg-n1EEV_nku-2nG3QRw) (psw: 36lm) to test on the DAVIS16 dataset, the mean-J and mean…
-
In the [newest published model](https://qurator-data.de/eynollah/2022-04-05/models_eynollah_renamed.tar.gz), all file paths have changed:
| 2021-04-25 | 2022-04-05 |
| --- | --- |
| model_3up_new…
-
The "Projects" tab should hold information about which projects are part of the latest ocrd_all release–and which are not. This information has to be provided within the `repos.json` (or a separate JS…
-
Thank you for your generous sharing!
I downloaded the CDD dataset mentioned in your paper, but found that the label of the CDD dataset was not binary map, as Issue in label dataset of 2018. Lebedev…
-
**Describe the bug**
Installing the latest version of lightwood throws ImportError. I guess the issue is related to the sciki-learn.
**To Reproduce**
Steps to reproduce the behavior:
1. Train mo…
-
Bei manchen Bildern sind die Linien zu dünn. Zu dünne Linien lassen sie verschwinden, wenn sie sich ausdehnen und erodieren.