{ "id": 186242, "title": "Research papers", "url": "/competitions/lyft-motion-prediction-autonomous-vehicles/discussion/186242", "author_name": "Carlos Souza", "post_date": "2020-09-23T18:16:38.704000", "votes": 44, "comment_count": 6, "views": 0, "content": "

Hey guys,
\nAfter a lot (really, a lot) of reading, finally I think I understand where's the best place to start. Over the past weeks, I've collected +40 recent papers in the space, which I'm sharing with you all in this public Mendeley group:

\n

https://www.mendeley.com/community/lyft-kaggle-research-papers/

\n

After scanning/quick reading them all, I believe these are the most promising worth implementing:

\n
    \n

  1. Multi-Agent Tensor Fusion for Contextual Trajectory Prediction

    2. \n

  2. Improving Movement Predictions of Traffic Actors in Bird’s-Eye View Models using GANs and Differentiable Trajectory Rasterization

    3. \n

  3. Uncertainty-aware short-term motion prediction of traffic actors for autonomous driving

    4. \n

  4. Rules of the road: Predicting driving behavior with a convolutional model of semantic interactions

    5. \n

  5. Multimodal trajectory predictions for autonomous driving using deep convolutional networks

    6. \n

  6. DESIRE: Distant Future Prediction in Dynamic Scenes with Interacting Agents

    7. \n

  7. Predicting Motion of Vulnerable Road Users using High-Definition Maps and Efficient ConvNets

    8. \n
\n

Is anyone implementing any of these? If so, I'm interested in teaming up with the purpose of studying & implementing those papers, to then apply them to the challenge.

\n

Cheers!

", "messages": [ { "id": 1024269, "postDate": "2020-09-23T18:16:38.703Z", "content": "

Hey guys,
\nAfter a lot (really, a lot) of reading, finally I think I understand where's the best place to start. Over the past weeks, I've collected +40 recent papers in the space, which I'm sharing with you all in this public Mendeley group:

\n

https://www.mendeley.com/community/lyft-kaggle-research-papers/

\n

After scanning/quick reading them all, I believe these are the most promising worth implementing:

\n
    \n

  1. Multi-Agent Tensor Fusion for Contextual Trajectory Prediction

    2. \n

  2. Improving Movement Predictions of Traffic Actors in Bird’s-Eye View Models using GANs and Differentiable Trajectory Rasterization

    3. \n

  3. Uncertainty-aware short-term motion prediction of traffic actors for autonomous driving

    4. \n

  4. Rules of the road: Predicting driving behavior with a convolutional model of semantic interactions

    5. \n

  5. Multimodal trajectory predictions for autonomous driving using deep convolutional networks

    6. \n

  6. DESIRE: Distant Future Prediction in Dynamic Scenes with Interacting Agents

    7. \n

  7. Predicting Motion of Vulnerable Road Users using High-Definition Maps and Efficient ConvNets

    8. \n
\n

Is anyone implementing any of these? If so, I'm interested in teaming up with the purpose of studying & implementing those papers, to then apply them to the challenge.

\n

Cheers!

", "rawMarkdown": "Hey guys,\nAfter a lot (really, a lot) of reading, finally I think I understand where's the best place to start. Over the past weeks, I've collected +40 recent papers in the space, which I'm sharing with you all in this public Mendeley group:\n\nhttps://www.mendeley.com/community/lyft-kaggle-research-papers/\n\nAfter scanning/quick reading them all, I believe these are the most promising worth implementing:\n\n1. [Multi-Agent Tensor Fusion for Contextual Trajectory Prediction](https://openaccess.thecvf.com/content_CVPR_2019/papers/Zhao_Multi-Agent_Tensor_Fusion_for_Contextual_Trajectory_Prediction_CVPR_2019_paper.pdf)\n\n2. [Improving Movement Predictions of Traffic Actors in Bird’s-Eye View Models using GANs and Differentiable Trajectory Rasterization](https://arxiv.org/pdf/2004.06247.pdf)\n\n3. [Uncertainty-aware short-term motion prediction of traffic actors for autonomous driving](https://openaccess.thecvf.com/content_WACV_2020/papers/Djuric_Uncertainty-aware_Short-term_Motion_Prediction_of_Traffic_Actors_for_Autonomous_Driving_WACV_2020_paper.pdf)\n\n4. [Rules of the road: Predicting driving behavior with a convolutional model of semantic interactions](https://openaccess.thecvf.com/content_CVPR_2019/papers/Hong_Rules_of_the_Road_Predicting_Driving_Behavior_With_a_Convolutional_CVPR_2019_paper.pdf)\n\n5. [Multimodal trajectory predictions for autonomous driving using deep convolutional networks](https://arxiv.org/pdf/1809.10732.pdf)\n\n6. [DESIRE: Distant Future Prediction in Dynamic Scenes with Interacting Agents](http://www.robots.ox.ac.uk/~tvg/publications/2017/CVPR17_DESIRE.pdf)\n\n7. [Predicting Motion of Vulnerable Road Users using High-Definition Maps and Efficient ConvNets](https://arxiv.org/pdf/1906.08469.pdf)\n\nIs anyone implementing any of these? If so, I'm interested in teaming up **with the purpose of studying & implementing those papers, to then apply them to the challenge.**\n\nCheers!", "votes": 44 }, { "id": 1029792, "postDate": "2020-09-28T06:17:45.573Z", "content": "

Nice to see the topics which are very interesting, thank you for sharing..!
\nIt would be really very great to share the learnings aswell.
\nIt would be good learning to implement as lot of challenging things exist.

", "rawMarkdown": "Nice to see the topics which are very interesting, thank you for sharing..!\nIt would be really very great to share the learnings aswell.\nIt would be good learning to implement as lot of challenging things exist." }, { "id": 1025121, "postDate": "2020-09-24T10:43:55.433Z", "content": "

Informative , Thanks for sharing @carlossouza

", "rawMarkdown": "Informative , Thanks for sharing @carlossouza " }, { "id": 1026089, "postDate": "2020-09-25T04:25:43.793Z", "rawMarkdown": "", "isDeleted": true }, { "id": 1026086, "postDate": "2020-09-25T04:24:00.360Z", "rawMarkdown": "", "isDeleted": true }, { "id": 1027616, "postDate": "2020-09-26T08:58:49.433Z", "content": "

Wow!! Thanks for sharing!

", "rawMarkdown": "Wow!! Thanks for sharing!" }, { "id": 1026593, "postDate": "2020-09-25T12:38:57.313Z", "content": "

Thanks for Sharing @carlossouza

", "rawMarkdown": "Thanks for Sharing @carlossouza" } ], "comments": [ { "id": 1029792, "author_name": "sravan jindham", "author_url": "", "post_date": "2020-09-28T06:17:45.573000", "content": "

Nice to see the topics which are very interesting, thank you for sharing..!
\nIt would be really very great to share the learnings aswell.
\nIt would be good learning to implement as lot of challenging things exist.

", "votes": 0, "replies": [] }, { "id": 1025121, "author_name": "Pinaki MIshra", "author_url": "", "post_date": "2020-09-24T10:43:55.433000", "content": "

Informative , Thanks for sharing @carlossouza

", "votes": 0, "replies": [] }, { "id": 1026089, "author_name": "", "author_url": "", "post_date": "2020-09-25T04:25:43.793000", "content": "", "votes": 0, "replies": [] }, { "id": 1026086, "author_name": "", "author_url": "", "post_date": "2020-09-25T04:24:00.360000", "content": "", "votes": 0, "replies": [] }, { "id": 1027616, "author_name": "Amritvir Singh", "author_url": "", "post_date": "2020-09-26T08:58:49.433000", "content": "

Wow!! Thanks for sharing!

", "votes": 0, "replies": [] }, { "id": 1026593, "author_name": "Umang Sachdev", "author_url": "", "post_date": "2020-09-25T12:38:57.313000", "content": "

Thanks for Sharing @carlossouza

", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "1024269": "Hey guys,\nAfter a lot (really, a lot) of reading, finally I think I understand where's the best place to start. Over the past weeks, I've collected +40 recent papers in the space, which I'm sharing with you all in this public Mendeley group:\n\nhttps://www.mendeley.com/community/lyft-kaggle-research-papers/\n\nAfter scanning/quick reading them all, I believe these are the most promising worth implementing:\n\n1. [Multi-Agent Tensor Fusion for Contextual Trajectory Prediction](https://openaccess.thecvf.com/content_CVPR_2019/papers/Zhao_Multi-Agent_Tensor_Fusion_for_Contextual_Trajectory_Prediction_CVPR_2019_paper.pdf)\n\n2. [Improving Movement Predictions of Traffic Actors in Bird’s-Eye View Models using GANs and Differentiable Trajectory Rasterization](https://arxiv.org/pdf/2004.06247.pdf)\n\n3. [Uncertainty-aware short-term motion prediction of traffic actors for autonomous driving](https://openaccess.thecvf.com/content_WACV_2020/papers/Djuric_Uncertainty-aware_Short-term_Motion_Prediction_of_Traffic_Actors_for_Autonomous_Driving_WACV_2020_paper.pdf)\n\n4. [Rules of the road: Predicting driving behavior with a convolutional model of semantic interactions](https://openaccess.thecvf.com/content_CVPR_2019/papers/Hong_Rules_of_the_Road_Predicting_Driving_Behavior_With_a_Convolutional_CVPR_2019_paper.pdf)\n\n5. [Multimodal trajectory predictions for autonomous driving using deep convolutional networks](https://arxiv.org/pdf/1809.10732.pdf)\n\n6. [DESIRE: Distant Future Prediction in Dynamic Scenes with Interacting Agents](http://www.robots.ox.ac.uk/~tvg/publications/2017/CVPR17_DESIRE.pdf)\n\n7. [Predicting Motion of Vulnerable Road Users using High-Definition Maps and Efficient ConvNets](https://arxiv.org/pdf/1906.08469.pdf)\n\nIs anyone implementing any of these? If so, I'm interested in teaming up **with the purpose of studying & implementing those papers, to then apply them to the challenge.**\n\nCheers!", "1029792": "Nice to see the topics which are very interesting, thank you for sharing..!\nIt would be really very great to share the learnings aswell.\nIt would be good learning to implement as lot of challenging things exist.", "1025121": "Informative , Thanks for sharing @carlossouza ", "1026089": "", "1026086": "", "1027616": "Wow!! Thanks for sharing!", "1026593": "Thanks for Sharing @carlossouza" } }