Відео

5:47

hey is it a real time robot riding there with the lidar or it was pre scanned and this is a model?

That is 3GB of storage and counting and not even talk of the computing, cameras over lidar

hey how was this scanned!

Fantastic work Charlie, the video is great! 🤩 (And the ORI too)

Boar was here

Good day. Do you create robots?

مهرجان الالوان النهاردة ياريت ت عم لووو يوم لالوان

Call me crazy but it's shit like this that makes me think the simulation theory has some weight to it. Sometimes I think we are just brains in jars having chemicals sent to us telling us that we are experiencing life when in reality we are just brains in jars. But again that's just a theory lol

Interesting work. The visualization of the gait trajectory through the latent space is very cool.

Hi sir I have small doubt what is the profession name of this job

Can you make this clean the hole bathroom please

Superb

💛 Great work!! I recommend using PromoSM!! ! It’s a handy tool to quickly grow your channel!

May I ask what is the name of the piano music in the last :), thank you :)?

834 vur.fyi

Smooth

Thank you for sharing.

Hi < Can you provide the code only the pedestrian part like what features u assigned to it?

Can it clean bathrooms?

I really like the Paul Newman way. I try to follow the same path. Thanks professor for sharing your thought.

very nice work !! I am very curious to know about the results of this approach in an environment which the model has not been trained on ?

Hi Great work! I have a question: Which exact Radar was used for this?

Can RADAR do dense reconstruction (mapping) of the environment? How accurate would that be?

Cool, I can imagine the day using radar to do localization, it will be soon

car.ai

It can not distinguish the static car and people, right? maybe it still decide by semantic class.

Paper : ori.ox.ac.uk/wp-content/uploads/2018/07/ICRA_paper_Cen_Newman.pdf

Is the paper available? If not what are the main improvements compared with last year approach: ori.ox.ac.uk/wp-content/uploads/2018/07/ICRA_paper_Cen_Newman.pdf

Great work, well done. I do recommend to mentiond the maximum,, minimum, and the error SD while showing slides related to the displacement- scan matching

3:19 , in the middle, a non-moving truck is masked. That truck can provide high quality features for VO though.

looks like it can only detect objects within ~20m range...

Just like humans, AI need clearer eyes

In which conference did you publish this research paper ? I think ICRA but could not find on IEEE

Hi do u have the open-source code?

It could also be useful for automated censorship of bystanders in a scene, useful for taking news footage.

Has any of this code been released or made public?

Hi, Could you explain more about the mechanism of the snake arm please? Is there any resources, Journals to discover more about it ? Thanks in advance :)

For those hungry to see more of OC Robotics' snake robots, OC Robotics does have their own youtube channel.

Hello, can I upload this video on my UA-cam channel vb.ai ?

Does it cross reference any types of maps?

Unusually ))

Awesome. Can you please share the learned models for us to try it in our dataset?

Impressive. I'll dive into the paper.

Great works, well done!!!

Any link to the paper ?

Can you give some info on the algorithms being used?

Hi, how did you get the ground truth? Manually label them to corresponding captured images? Or calculated using inertial sensors? Cheers.

Interesting solution to match on the yz plane instead of the normal xy. Then you "cover" all 3D even though you use a 2D LIDAR. Is this stable to rotational differences when looking for loop closure? Lets say you enter an intersection from a different direction?

awesome stuff guys, is this something new beyond your IJRR paper in the LTA issue last year?