Hao Li
Hao Li
In pretrain procedure, only *foreground* and *background* is unsed
Because both freemask and self-sup training are total the same, except for the dataset. The annotations of self-sup cames from the predictions of freemask training. It's general in self-sup training.
Same problem. Wish to be responded.
It seems the model aims to detect the background rather than the object. How's your mAP result? Is it also strange? ## FYI I've tested the model in COCO dataset,...
Something must happen in your code. I've tested in your demo image, the result is shown below:  My vscode config code: ```json { "env": {"PYTHONPATH" : "${workspaceRoot}"}, "name": "solov2-coco:demo",...
Yes, just delete it. I was testing low-data dataset in that moment.
Again, it would be useful if you can provide your training logs. Because the demo you train on official project(WXlong/SOLO) is also strange, it contains overlap parts. Yet the **Matrix...
A personal question, do you speak Chinese? We can talk in chinese. ## DEMO code The **demo.py** directly comes from the official mmdetection project. ```python ''' Author: your name Date:...
## Results I'm using 2.19.0 exactly, and the results show no GAP between my implementation and the official version. ## Tips The official checkpoint file should be modified before you...