Klein.Peng comments

Results 5 comments of


                                            Klein.Peng

Question on the Loss function

1. total_loss = 1/N sum_for_every_image(image_loss) 2. If the IOU between prediction box and ground truth box is biger than 0.6, I think it's a right cell. Otherwise, it contribute to...

could you list the finished function of your code?

I can train successfully. Because there is no pre-train weights used, I spent a lot of time training the model, and it's accuracy is lower than darknet YOLOv2 obviously. I...

could you list the finished function of your code?

I have the code to compute mAP and dimension cluster,but I am glad to have a partner to complete my project

could you list the finished function of your code?

Now the code is still a simple prototype, I will optimize it step by step. Finally I will also write a guide.

preprocess_data with tensorflow is slow

1. The data got from TFRecord is tensor. 2.The way you evaluate the speed is not appropriate for my code. It's more reasonable that we evaluate the speed with vectorization.