search

markshih91 / mobilenet_v2_ssdlite_keras

A keras version of real-time object detection network : mobilenet_v2_ssdlite
80 stars 40 forks source link

How sub_sample the weights? #3

Open zhoujinhai opened 5 years ago

zhoujinhai commented 5 years ago

@markshih91, I want to training the model on my dataset that have two classess, how to sub-sample the weight tensors of all the classification layers?

DubiousCactus commented 5 years ago

The original writer of this SSD implementation has a very nice tutorial for that: https://github.com/pierluigiferrari/ssd_keras/blob/master/weight_sampling_tutorial.ipynb

zhoujinhai commented 5 years ago

yes, I have tried it, and change the layers name from conv4_3_norm_mbox_conf to ssd_cls1conv2, but it show that bias:0 does not exist after running.

DubiousCactus commented 5 years ago

Well I can't help you, I haven't tried it myself...

zhoujinhai commented 5 years ago

Ok Thanks

发自我的iPhone

------------------ Original ------------------ From: Théo Morales <notifications@github.com> Date: Wed,Sep 18,2019 0:20 AM To: markshih91/mobilenet_v2_ssdlite_keras <mobilenet_v2_ssdlite_keras@noreply.github.com> Cc: zhoujinhai <932853432@qq.com>, Author <author@noreply.github.com> Subject: Re: [markshih91/mobilenet_v2_ssdlite_keras] How sub_sample the weights? (#3)

Well I can't help you, I haven't tried it myself...

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.

liguilan1227 commented 4 years ago

yes, I have tried it, and change the layers name from conv4_3_norm_mbox_conf to ssd_cls1conv2, but it show that bias:0 does not exist after running.

I met the problem,too. Have you solve it? Please help me.