-
### Expected Behavior
I expect nodes and lines and groups to scale with each other when I zoom in and out
### Actual Behavior
Since the latest git pull + restart comfy (which also updates front end…
-
```(nerfstudio) C:\nerfstudio\feature-splatting>ns-train feature-splatting --data data/nerfstudio/poster
C:\Users\local_account\anaconda3\envs\nerfstudio\lib\site-packages\nerfstudio\utils\misc.py:18…
-
### Frontend Version
1.2.28
### Expected Behavior
Nigh-instant startup of UI after clicking the URL from the console after comfy launch.
### Actual Behavior
I see a spinning wheel for quite a lo…
-
I trained LightGlue on the homographies and then moved on to finetune against MegaDepth.
In the megadepth training script., just at the very start, when loading the model state, the training script t…
-
After downloading comfyUI portable (FRESH INSTALLATION), going to python embedded directoty and removing Torch 2.4.1, then installing Torch 2.4.0, then trying to install the very difficult to install …
-
{"logger": "uvicorn.error", "timestamp": "2024-09-20T14:44:45.950538Z", "severity": "INFO", "message": "Started server process [7]"}
{"logger": "uvicorn.error", "timestamp": "2024-09-20T14:44:45.9506…
-
Hi, I use the `convert_to_pytorch` code to convert my TF model. The results are very different.
1. TF version
-
Hello, I have been trying to use lightglue to extract features from 2 consecutive frames obtained from a stereo camera (real time camera stream). However everytime I try to extract features between tw…
-
How can I reconstruct one sfm model by Pinpole in the deep-image-matching?