Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Several key issues on map change labels #11

Open
qiaozhijian opened this issue Jul 18, 2022 · 3 comments
Open

Several key issues on map change labels #11

qiaozhijian opened this issue Jul 18, 2022 · 3 comments

Comments

@qiaozhijian
Copy link

qiaozhijian commented Jul 18, 2022

I find some inconsistency (from my perspective) of map change labels. This may mean that the labeling of map changes is a very custom work, related to each person's judgment criteria and map elements of interest. Is it necessary to further improve the labels of map change?

There are two examples: from left to right are the front view, the rendered map, and the semantic segmentation.

First, log 01bb304d7bd835f8bbef7086b688e35e__Summer_2019 at timestamp 315968879949927215. This place has been labeled as no change, but you can see that the lane lines are noticeably faded.
179478103-26a99397-bcc7-4b1f-b2b8-0514a80bdeb0

Second, log 1SKq8bGqPRZq6PUbFgTwVLyMZEwtiD2S__Summer_2020 at timestamp 315966949549927209. Maybe because the annotator didn't pay attention to the stop line and didn't draw it into the map, but the stop line was segmented so that the map was considered to have changed.

179478526-f98f4bb2-fd00-41a4-8e6c-ab75d02d7360

@qiaozhijian
Copy link
Author

There is another example of BEV in 2S1xkgWyVhoa5zwPLz9tl2MAG0lMxj6i on 315980583459840000. Obviously, the sidewalk has disappeared, but there are still obvious sidewalk markings on the map, and this case is labeled as "no change".

Filename: 2S1xkgWyVhoa5zwPLz9tl2MAG0lMxj6i__Autumn_2020_315980583459840000_rgb_interpTruelinear_projmethodray_tracing.jpg

2S1xkgWyVhoa5zwPLz9tl2MAG0lMxj6i__Autumn_2020_315980583459840000_rgb_interpTruelinear_projmethodray_tracing

Filename: 2S1xkgWyVhoa5zwPLz9tl2MAG0lMxj6i__Autumn_2020_315980583459840000_vectormap.jpg

2S1xkgWyVhoa5zwPLz9tl2MAG0lMxj6i__Autumn_2020_315980583459840000_vectormap

@johnwlambert
Copy link
Owner

Hi @qiaozhijian, thanks for your note. There are expected to be a few minor errors from human labeling, as more than 10,000 frames are labeled and evaluated. However, let me take a closer look at these. In the last comment, did you mean "crosswalk" instead of "sidewalk"?

@qiaozhijian
Copy link
Author

Hi @qiaozhijian, thanks for your note. There are expected to be a few minor errors from human labeling, as more than 10,000 frames are labeled and evaluated. However, let me take a closer look at these. In the last comment, did you mean "crosswalk" instead of "sidewalk"?

Thank you for your reply. I'm sorry I made a mistake. It's "crosswalk".

Looking forward to this dataset being better. Your work has really contributed to the research community, thank you very much!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants