search

iitmcvg / CNN_SLAM

CNN SLAM implementation of https://arxiv.org/abs/1704.03489
http://iitmcvg.github.io/projects/CNN_SLAM
198 stars 50 forks source link

does this work? #1

Open graphic-zhang opened 5 years ago

graphic-zhang commented 5 years ago

it seems not complete?

abdur4373 commented 5 years ago

same here I want to implement CNN SLAM but could not find the code. Can someone help me out.

graphic-zhang commented 5 years ago

same here I want to implement CNN SLAM but could not find the code. Can someone help me out.

if you find , please tell me ~

abdur4373 commented 5 years ago

Do you have any alternative or any idea because I am totally new in this field.

graphic-zhang commented 5 years ago

Do you have any alternative or any idea because I am totally new in this field.

I am finding too, actually, it does two things new, fuse depth prediction and depth map, semantic segment and loop closure, you can try respectively

abdur4373 commented 5 years ago

I have done depth prediction through paper of laina "Deeper depth prediction ...." have implemented LSD SLAM as well, Now I do not know how to proceed and how to use LSD SLAM in CNN SLAM. I am doing all this in ros kinetic and delth prediction in pytorch. If you could guide me now how to proceed. My email is abdur4373@gmail.com.

graphic-zhang commented 5 years ago

I have done depth prediction through paper of laina "Deeper depth prediction ...." have implemented LSD SLAM as well, Now I do not know how to proceed and how to use LSD SLAM in CNN SLAM. I am doing all this in ros kinetic and delth prediction in pytorch. If you could guide me now how to proceed. My email is abdur4373@gmail.com.

of course you can communicate with me, but I have worked now, most of time I am very busy,so you can write the email about" what you did and what the problem" to zhangxiang941211@gmail.com

Foexy commented 5 years ago

Can somebody help me with CNN-SLAM?

viet2411 commented 5 years ago

Has anyone successfully using this code? Can anybody give me simple guide?