You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Thanks for ur excellent work! I tried the same image and the same backbone on Hugging Face and on my computer separately, but I got different results. It's very clear that the results on my computer cannot recognize water.
The text was updated successfully, but these errors were encountered:
How should I do to improve the performance? The above results are from using DINAT as the backbone, trained on the COCO dataset, for making predictions.
+1
I'm experiencing the same issue. The Hugging Face demo performs significantly better. I compared it locally with the same model and the same images without modifying anything, yet the performance discrepancy remains.
Thanks for ur excellent work! I tried the same image and the same backbone on Hugging Face and on my computer separately, but I got different results. It's very clear that the results on my computer cannot recognize water.
The text was updated successfully, but these errors were encountered: