{ "id": 178319, "title": "@Host : Clarification regarding 4th Agent in Data Format description", "url": "/competitions/lyft-motion-prediction-autonomous-vehicles/discussion/178319", "author_name": "", "post_date": "2020-08-29T14:26:09.308778900Z", "votes": 8, "comment_count": 3, "views": 0, "content": "

According to the description in dataset (Please click Data Tab of competition page to find it). There should be 4 out of 17 possible agents should be present. But if we look at data three of them are car, pedestrian and cyclists. And the fourth one has the tag UNKNOWN. \"\"

", "messages": [ { "id": "990332", "postDate": "08/29/2020 14:26:09", "content": "

According to the description in dataset (Please click Data Tab of competition page to find it). There should be 4 out of 17 possible agents should be present. But if we look at data three of them are car, pedestrian and cyclists. And the fourth one has the tag UNKNOWN. \"\"

", "rawMarkdown": "According to the description in dataset (Please click Data Tab of competition page to find it). There should be 4 out of 17 possible agents should be present. But if we look at data three of them are car, pedestrian and cyclists. And the fourth one has the tag UNKNOWN. ![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F344232%2F4c6bbdccc038b92e3610c2da4b44dcad%2FKagg.png?generation=1598711613397575&alt=media)", "votes": null }, { "id": "990640", "postDate": "08/29/2020 18:52:28", "content": "

I like the \"don't care!\" label. There are even two of them, but who cares :)

\n

Also I am perplexed what is the difference between motorcyclist and pedestrian, given that there is a separate motorcycle type. Probably can be identified by a helmet and generally reckless behavior.

", "rawMarkdown": "I like the \"don't care!\" label. There are even two of them, but who cares :)\n\nAlso I am perplexed what is the difference between motorcyclist and pedestrian, given that there is a separate motorcycle type. Probably can be identified by a helmet and generally reckless behavior.", "votes": null }, { "id": "990759", "postDate": "08/29/2020 20:29:13", "content": "

@zaharch I think the american definition right now for a motorcyclist is motor drive two wheeler on the road. And pedestrian is everything on the walk path. So scooter capital cities like berlin or california should consider a scooter driver a running pedestrian 😂 (just joking)

", "rawMarkdown": "zaharch I think the american definition right now for a motorcyclist is motor drive two wheeler on the road. And pedestrian is everything on the walk path. So scooter capital cities like berlin or california should consider a scooter driver a running pedestrian 😂 (just joking)", "votes": null }, { "id": "994249", "postDate": "09/01/2020 13:39:58", "content": "

Hi @deepakrajpurushothaman, the 4th label (UNKNOWN) is for everything the perception stack is not able to identify. You won't be asked to predict for those agents (they are not included in the test mask). We may want to update the page to specify that :)

", "rawMarkdown": "Hi @deepakrajpurushothaman, the 4th label (UNKNOWN) is for everything the perception stack is not able to identify. You won't be asked to predict for those agents (they are not included in the test mask). We may want to update the page to specify that :)", "votes": null } ], "comments": [ { "id": 990640, "author_name": "zaharch", "author_url": "", "post_date": "08/29/2020 18:52:28", "content": "

I like the \"don't care!\" label. There are even two of them, but who cares :)

\n

Also I am perplexed what is the difference between motorcyclist and pedestrian, given that there is a separate motorcycle type. Probably can be identified by a helmet and generally reckless behavior.

", "votes": null, "replies": [ { "id": 990759, "author_name": "deepakrajpurushothaman", "author_url": "", "post_date": "08/29/2020 20:29:13", "content": "

@zaharch I think the american definition right now for a motorcyclist is motor drive two wheeler on the road. And pedestrian is everything on the walk path. So scooter capital cities like berlin or california should consider a scooter driver a running pedestrian 😂 (just joking)

", "votes": null, "replies": [] } ] }, { "id": 994249, "author_name": "lucabergamini", "author_url": "", "post_date": "09/01/2020 13:39:58", "content": "

Hi @deepakrajpurushothaman, the 4th label (UNKNOWN) is for everything the perception stack is not able to identify. You won't be asked to predict for those agents (they are not included in the test mask). We may want to update the page to specify that :)

", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "990332": "According to the description in dataset (Please click Data Tab of competition page to find it). There should be 4 out of 17 possible agents should be present. But if we look at data three of them are car, pedestrian and cyclists. And the fourth one has the tag UNKNOWN. ![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F344232%2F4c6bbdccc038b92e3610c2da4b44dcad%2FKagg.png?generation=1598711613397575&alt=media)", "990640": "I like the \"don't care!\" label. There are even two of them, but who cares :)\n\nAlso I am perplexed what is the difference between motorcyclist and pedestrian, given that there is a separate motorcycle type. Probably can be identified by a helmet and generally reckless behavior.", "990759": "zaharch I think the american definition right now for a motorcyclist is motor drive two wheeler on the road. And pedestrian is everything on the walk path. So scooter capital cities like berlin or california should consider a scooter driver a running pedestrian 😂 (just joking)", "994249": "Hi @deepakrajpurushothaman, the 4th label (UNKNOWN) is for everything the perception stack is not able to identify. You won't be asked to predict for those agents (they are not included in the test mask). We may want to update the page to specify that :)" }, "source": "meta" }