luissen

Peking university interested in SISR, domain adaptation reid and object detection.

Results 6 comments of


                                            luissen

Problem with importing ModulatedDeformConvPack

How to solve it?

About the source_train.py

I just delete it...... orz

Training time

it is right. Your 16000 is 200 iterations, so the time is double.

Training time

@wsqat you can change the batchsize

can you tell me why it is 23 for k in range(23):

because the last two layers of VGG should be treated alone

[Bug] Yoloact does not conver to onnx successfully using mmdeploy

> I think you can use yolov8 instead of MM, the segmentation part of yolov8 is based on yoloact. yolov8 not work, I tried.