Deep Knowledge

Cut the model to save time

5

the Resnet has 4 layers, so if you want to use only 3 layers (1-3), you can cut the model, this will save the time I tried it in my...

Speed up the PaDim

2

I have optimized the inference time, so now it is about 50 fps, above to that the model can be exported to onnx and openvino. It also exceed the time...

Target and input must have the same number of elements. target nelement (20) != input nelement (3380)

2

I changed the batchsize = 20, in line 117, disc_loss_real = L_adv_disc(D_real, label_real.detach()) i got this error, Target and input must have the same number of elements. target nelement (20)...

Optimize version

3

I'm about to publish my Optimized version. The inference time of onnx on gpu 4070 is about 5 ms. It has been converted to several format, including onnx, tf, hailo...