{
  "id": 192496,
  "title": "Traffic light ahead of agent",
  "url": "/competitions/lyft-motion-prediction-autonomous-vehicles/discussion/192496",
  "author_name": "",
  "post_date": "2020-10-21T20:41:48.255695800Z",
  "votes": 3,
  "comment_count": 4,
  "views": 0,
  "content": "<p>Hello All,<br>\nI have a small question. Do anyone know how to get info about the traffic light ahead of agent (for a given timestamp)?</p>\n<p>Thanks,<br>\nBoris</p>",
  "messages": [
    {
      "id": "1056567",
      "postDate": "10/21/2020 20:41:48",
      "content": "<p>Hello All,<br>\nI have a small question. Do anyone know how to get info about the traffic light ahead of agent (for a given timestamp)?</p>\n<p>Thanks,<br>\nBoris</p>",
      "rawMarkdown": "Hello All,\nI have a small question. Do anyone know how to get info about the traffic light ahead of agent (for a given timestamp)?\n\nThanks,\nBoris",
      "votes": null
    },
    {
      "id": "1057104",
      "postDate": "10/22/2020 11:16:28",
      "content": "<p>You can get the traffic light faces from the zarr. I would look into the source of the semantic rasterizer to see, how the traffic light data is extracted and used.</p>",
      "rawMarkdown": "You can get the traffic light faces from the zarr. I would look into the source of the semantic rasterizer to see, how the traffic light data is extracted and used.",
      "votes": null
    },
    {
      "id": "1057237",
      "postDate": "10/22/2020 13:51:52",
      "content": "<p>Thank you, Marek!</p>",
      "rawMarkdown": "Thank you, Marek!",
      "votes": null
    },
    {
      "id": "1057724",
      "postDate": "10/22/2020 23:33:39",
      "content": "<p>These lines in semantic rasterizer will give you traffic light info.</p>\n<p><code>for tl_id in lane_tl_ids.intersection(active_tl_ids):\n                if self.proto_API.is_traffic_face_colour(tl_id, \"red\"):\n                    lane_type = \"red\"\n                elif self.proto_API.is_traffic_face_colour(tl_id, \"green\"):\n                    lane_type = \"green\"\n                elif self.proto_API.is_traffic_face_colour(tl_id, \"yellow\"):\n                    lane_type = \"yellow\"\n</code></p>",
      "rawMarkdown": "These lines in semantic rasterizer will give you traffic light info.\n\n`for tl_id in lane_tl_ids.intersection(active_tl_ids):\n                if self.proto_API.is_traffic_face_colour(tl_id, \"red\"):\n                    lane_type = \"red\"\n                elif self.proto_API.is_traffic_face_colour(tl_id, \"green\"):\n                    lane_type = \"green\"\n                elif self.proto_API.is_traffic_face_colour(tl_id, \"yellow\"):\n                    lane_type = \"yellow\"\n`",
      "votes": null
    },
    {
      "id": "1058386",
      "postDate": "10/23/2020 16:32:42",
      "content": "<p>Sujayd, thanks a lot!</p>",
      "rawMarkdown": "Sujayd, thanks a lot!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1057104,
      "author_name": "marekwyborski",
      "author_url": "",
      "post_date": "10/22/2020 11:16:28",
      "content": "<p>You can get the traffic light faces from the zarr. I would look into the source of the semantic rasterizer to see, how the traffic light data is extracted and used.</p>",
      "votes": null,
      "replies": [
        {
          "id": 1057237,
          "author_name": "polishch",
          "author_url": "",
          "post_date": "10/22/2020 13:51:52",
          "content": "<p>Thank you, Marek!</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1057724,
      "author_name": "sujaydkhandekar",
      "author_url": "",
      "post_date": "10/22/2020 23:33:39",
      "content": "<p>These lines in semantic rasterizer will give you traffic light info.</p>\n<p><code>for tl_id in lane_tl_ids.intersection(active_tl_ids):\n                if self.proto_API.is_traffic_face_colour(tl_id, \"red\"):\n                    lane_type = \"red\"\n                elif self.proto_API.is_traffic_face_colour(tl_id, \"green\"):\n                    lane_type = \"green\"\n                elif self.proto_API.is_traffic_face_colour(tl_id, \"yellow\"):\n                    lane_type = \"yellow\"\n</code></p>",
      "votes": null,
      "replies": [
        {
          "id": 1058386,
          "author_name": "polishch",
          "author_url": "",
          "post_date": "10/23/2020 16:32:42",
          "content": "<p>Sujayd, thanks a lot!</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1056567": "Hello All,\nI have a small question. Do anyone know how to get info about the traffic light ahead of agent (for a given timestamp)?\n\nThanks,\nBoris",
    "1057104": "You can get the traffic light faces from the zarr. I would look into the source of the semantic rasterizer to see, how the traffic light data is extracted and used.",
    "1057237": "Thank you, Marek!",
    "1057724": "These lines in semantic rasterizer will give you traffic light info.\n\n`for tl_id in lane_tl_ids.intersection(active_tl_ids):\n                if self.proto_API.is_traffic_face_colour(tl_id, \"red\"):\n                    lane_type = \"red\"\n                elif self.proto_API.is_traffic_face_colour(tl_id, \"green\"):\n                    lane_type = \"green\"\n                elif self.proto_API.is_traffic_face_colour(tl_id, \"yellow\"):\n                    lane_type = \"yellow\"\n`",
    "1058386": "Sujayd, thanks a lot!"
  },
  "source": "meta"
}